oral health literacy and oral hygiene habits in a kentucky
TRANSCRIPT
Walden UniversityScholarWorks
Walden Dissertations and Doctoral Studies Walden Dissertations and Doctoral StudiesCollection
2019
Oral Health Literacy and Oral Hygiene Habits in aKentucky Appalachian CommunityKatie D. SchillWalden University
Follow this and additional works at: https://scholarworks.waldenu.edu/dissertations
Part of the Dentistry Commons, Epidemiology Commons, and the Public Health Education andPromotion Commons
This Dissertation is brought to you for free and open access by the Walden Dissertations and Doctoral Studies Collection at ScholarWorks. It has beenaccepted for inclusion in Walden Dissertations and Doctoral Studies by an authorized administrator of ScholarWorks. For more information, pleasecontact [email protected].
Walden University
College of Health Sciences
This is to certify that the doctoral dissertation by
Katie Schill
has been found to be complete and satisfactory in all respects,
and that any and all revisions required by
the review committee have been made.
Review Committee
Dr. Lori Dewald, Committee Chairperson, Public Health Faculty
Dr. Hebatullah Tawfik, Committee Member, Public Health Faculty
Dr. Kai Stewart, University Reviewer, Public Health Faculty
Chief Academic Officer
Eric Riedel, Ph.D.
Walden University
2019
Abstract
Oral Health Literacy and Oral Hygiene Habits in a Kentucky Appalachian Community
by
Katie Schill
BA, Eastern Kentucky University, 2009
Dissertation Submitted in Partial Fulfillment
of the Requirements for the Degree of
Doctor of Philosophy in Public Health, Epidemiology Specialization
Walden University
May 2019
Abstract
This study sought to identify the level of oral health literacy held by people who live in
transitional and distressed Kentucky Appalachian areas and if this effects how often they
are using oral hygiene techniques. Data were also collected to describe the attitudes
Kentucky Appalachian adults hold toward oral hygiene and oral health status. Current
documentation shows that poor oral health remains a public health threat in this
population despite efforts such as school-based sealant programs and increased dental
insurance coverage. This study followed a quantitative design and 99 participants were
polled using a survey specifically developed for this study’s use. Composite median
scores and Spearman’s correlation values established the existence of a low oral health
literacy level across the participant pool, an also documented that oral hygiene techniques
are not used in frequencies recommended for proper oral health. A poor self-efficacy
towards the ability to utilize these techniques properly was also identified. Using the
Mann-Whitney U test, responses were compared based on county designation and few
significant differences were found. These findings show that oral health status and related
beliefs are similar across the region and not just isolated to the economically poorest
areas as the currently available literature suggests. Applying the health belief model it is
predicted that Kentucky Appalachians are unlikely to adopt proper oral hygiene habits
until their self-efficacy is improved. A recommendation of this study is that public health
officials should promote personal control when designing public health programs geared
towards improving the oral health status of this population. To do so would introduce a
positive social change in that people with good oral health are less likely to experience
the pain, malnutrition, and negative social stigma that is associated with poor oral health.
Oral Health Literacy and Oral Hygiene Habits in a Kentucky Appalachian Community
by
Katie Schill
BA, Eastern Kentucky University, 2009
Dissertation Submitted in Partial Fulfillment
of the Requirements for the Degree of
Doctor of Philosophy in Public Health, Epidemiology Specialization
Walden University
May 2019
Acknowledgements
To Karl H. Schill, Eva McKee RN, and Ellyn Schill PharmD, even though you
guys have to love me it has been great having all of your support! I will expect this to
continue when the student loan payments start!
To Carrie M. Oakley RN, for planting the idea of an Epi degree into my brain. I
owe you one, Chickie.
To Dianne Richeson CRC, for always being my cheerleader.
To Linda S. Rice RN, CRC, for the support, the help, the advice, the willingness
to be a sounding board, and for listening to all the complaining.
To Lori N. Sizemore, for always telling me that I could do anything and be
anything I wanted.
To Seth M. Neil: you know what you have put up with. You are a saint among
men.
To Dr. Heba Tawfik for being willing to read through copious amounts of Word
documents and for always challenging me to do my absolute best.
And to Dr. Lori Dewald for all of the guidance, instruction, advice, and support
throughout this process. This dissertation simply would not exist without you. Thank you
for always being so free in sharing your expertise and experiences with me: I have truly
been the most fortunate of PhD students!
i
Table of Contents
Chapter 1: Introduction to the Study ....................................................................................1
Introduction ....................................................................................................................1
Background ....................................................................................................................3
Research Literature Summary ................................................................................. 3
Gap in Knowledge .................................................................................................. 4
Need for Study ........................................................................................................ 5
Problem Statement .........................................................................................................7
Purpose of Study ............................................................................................................9
Independent Variable .............................................................................................. 9
Dependent Variables ............................................................................................. 10
Predictor Variable ................................................................................................. 11
Research Questions and Descriptive Items ..................................................................11
Theoretical Framework ................................................................................................13
Nature of the Study ......................................................................................................15
Definitions....................................................................................................................15
Assumptions .................................................................................................................16
Scope and Delimitations ..............................................................................................17
Limitations ...................................................................................................................18
Significance..................................................................................................................19
Summary ......................................................................................................................20
Chapter 2: Literature Review .............................................................................................22
ii
Introduction ..................................................................................................................22
Literature Search Strategy............................................................................................23
Theoretical Foundation ................................................................................................26
Literature Review Related to Key Variables and/or Concepts ....................................33
Kentucky Appalachian Region ............................................................................. 33
Independent Variable: Oral Health Literacy ......................................................... 37
Dependent Variable: Oral Hygiene Methods ........................................................ 40
Dependent Variable: Perceived Control of Oral Health ....................................... 41
Dependent Variable: Perceived Importance of Oral Hygiene .............................. 42
Predictor Variable: Dental Decay ......................................................................... 44
Summary and Conclusion ............................................................................................44
Chapter 3: Research Method ..............................................................................................47
Introduction ..................................................................................................................47
Research Design and Rationale ...................................................................................47
Methodology ................................................................................................................50
Population ............................................................................................................. 50
Sampling Procedures ............................................................................................ 52
Recruitment Procedures ........................................................................................ 52
Sample Size ........................................................................................................... 53
Participation Procedures ....................................................................................... 55
Data Collection ..................................................................................................... 56
Instrumentation ............................................................................................................57
iii
Reliability .............................................................................................................. 58
Validity ................................................................................................................. 58
Operationalization ................................................................................................. 59
Data Analysis Plan ................................................................................................ 60
Analysis Plan ........................................................................................................ 62
Threats to Validity .......................................................................................................67
External Validity - Threats.................................................................................... 67
Internal Validity - Threats ..................................................................................... 69
Construct – Threats ............................................................................................... 71
Statistical Conclusion Validity – Threats ............................................................. 72
Ethical Procedures .......................................................................................................73
Institutional Permissions ....................................................................................... 73
Ethical Concerns ................................................................................................... 73
Data Type and Storage .......................................................................................... 74
Compensation ....................................................................................................... 75
Summary ......................................................................................................................75
Chapter 4: Data Analysis ...................................................................................................76
Introduction ..................................................................................................................76
Data Collection ............................................................................................................79
Results 82
Research Questions and Descriptive Items Data Analysis ..........................................84
Research Question 1 ............................................................................................. 86
iv
Research Question 2 ............................................................................................. 90
Research Question 3 ............................................................................................. 91
Research Question 4 ............................................................................................. 92
Descriptive Information ........................................................................................ 96
Summary ....................................................................................................................106
Chapter 5: Final Interpretation .........................................................................................117
Introduction ................................................................................................................117
Interpretation of the Findings.....................................................................................117
Limitations of the Study.............................................................................................126
Recommendations ......................................................................................................129
Implications................................................................................................................134
Conclusion .................................................................................................................138
References ..................................................................................................................140
Appendix ..........................................................................................................................164
v
List of Tables
Table 1 Databases and search engines used in this study .......................................... 23
Table 2 Cronbach’s alpha scores for Research Question 1........................................ 89
Table 3 Cronbach’s alpha scores for Research Question 2........................................ 91
Table 4 Reported usage of oral hygiene techniques for adults and for children ....... 92
Table 5 Mann-Whitney U test results: Retain null .................................................... 95
Table 6 Mann-Whitney U test results: Reject null..................................................... 96
Table 7 Descriptive Item 1 Results ............................................................................ 98
Table 8 Descriptive Item 2 Results ............................................................................ 99
Table 9 Descriptive Item 3 Results ............................................................................ 101
Table 10 Descriptive Item 4 Results ......................................................................... 102
Table 11 Descriptive Item 5 Results ......................................................................... 103
Table 12 Descriptive Item 6 Results ......................................................................... 104
Table 13 Descriptive Item 7 Results ......................................................................... 105
Table 14 Descriptive Item 8 IOR Percentiles per survey item ................................. 111
Table 15 Descriptive Item 8 response frequencies ................................................... 112
1
Chapter 1: Introduction to the Study
Introduction
Traditionally the Appalachian population has demonstrated a lower level of
functional health literacy when compared to their non-Appalachian counterparts (Hutson,
Dorgan, Phillips, & Behringer, 2007; Ludke, Obermiller, Jacobson, Shaw, & Wells,
2006). Along with this lower health literacy level, the Appalachian population
experiences a higher rate of health disparities across many chronic disease conditions
(Ludke & Obermiller, 2016). These health disparities also include a substandard level of
oral health, which is a trend demonstrated by Kentucky’s population that has elevated
rates of dental decay especially concentrated in its vast Appalachian population (Dawkins
et al., 2013; Ludke & Obermiller, 2016; Saman, Johnson, Arevalo, & Odoi, 2011).
Poor oral health not only causes a variety of dental diseases but is also associated
as a risk-factor for many other chronic diseases as well as being recognized as a cofactor
by increasing the chance that an existing chronic disease will worsen in its intensity
(Cullinan, Ford, & Seymour, 2009; National Heart, Lung, and Blood Institute, 2016).
Along with influencing physical conditions, poor oral health has also been established as
negatively affecting mental health in terms of such aspects as self-esteem, school and job
performance, and social interactions (Sischo & Broder, 2011). Despite high rates of
fluoridated water supplies and school-based sealant programs, Kentucky is continuously
ranked among the highest in the nation for poor oral health indices (Saman et al., 2011).
The 2012 Behavioral Risk Factor Surveillance System (BRFSS) indicated that 51.5% of
Kentucky adults over the age of 65 years have had six or more permanent teeth removed,
2
compared to the U.S. national rate of 39.6% (Oral Health in Kentucky Technical Report,
2016). The BRFSS also demonstrated that tooth loss in Kentucky adults displayed
regional based trends. Kentuckians located in the eastern-most Appalachian areas
reported above state-level rates of tooth loss. Dental disease is not just found in
Appalachian adults: children also display dental complications, with Kentucky again
ranked among the highest in the nation for the prevalence of childhood dental caries.
These data indicate a cycle of poor dental health that is being passed from generation to
generation.
Currently accepted and documented reasons for the increased rates of dental
decay in Kentucky Appalachians include lower access to dental health professionals or
clinics, as well as a decreased access to dental insurance as compared to non-
Appalachians (Dawkins et al., 2013; Krause, May, Lane, Cossman, & Konrad, 2012).
However, little to no documentation can be found that examines the oral hygiene habits
of Kentucky Appalachian adults and their children. Oral hygiene habits such as brushing
teeth and flossing are documented as being an integral part in preventing most cases of
dental decay and so are very important to the overall status of oral health (Van der
Weijden & Slot, 2010).
This study will evaluate for a relationship between level of oral health literacy and
the frequency of oral hygiene techniques used by caregiver’s and their dependent children
located in a Kentucky Appalachian community. If such a relationship can be established,
this could indicate that specialized educational efforts featuring oral hygiene habits may
assist in improving the oral health status of this population. The perceptions and attitudes
3
held towards good oral health, as well as poor oral health, will also be measured. By
documenting the beliefs held by caregiver’s towards oral health, this could provide
insight into an area that needs targeting by public health efforts to improve the local
understanding of dental disease and its associated health risks.
Background
Research Literature Summary
The concept of health literacy is used to describe how well a person processes and
understands health related information (Nutbeam, 2008). Having a poor level of health
literacy is often indicative of having an overall poor health status and poorer health
outcomes, while those who demonstrate higher levels of health literacy usually have
more positive health statuses and outcomes (Nutbeam, 2008). Nielsen-Bohlman, Panzer,
and Kindig acknowledged that this relationship between health literacy level and overall
health status has now become a recognized and essential component of health care (as
cited by Vann, Lee, Baker, & Divaris, 2010, p.1395). Similar to health literacy, the term
oral health literacy has recently developed to describe a person’s ability to obtain,
process, and understand information that pertains to dental health and so is considered to
be a part of a person’s overall level of health literacy (Vann et al., 2010).
Despite the acceptance of health literacy as a part of a successful health care plan,
an estimated 80 million adults across the United States are believed to have limited health
literacy levels (Berkman, Sheriadan, Donahue, Halpern, & Crotty, 2011). Having low
health literacy may place these people at an increased risk of poor health status and
outcomes (Berkman et al., 2011). Certain groups and populations demonstrate higher
4
rates of low health literacy. These groups include those with a low socioeconomic status
and those who have less than a high school education and both of these factors are found
in abundance across the Kentucky Appalachian region (Berkman et al., 2011; Borak,
Salipante-Zaidel, Slade, & Fields, 2012; Elam, 2002). In addition to these shared
indicative factors, Ludke et al. (2006) documented that Kentucky Appalachians are
significantly more likely to have a lower level of health literacy when compared to their
non-Appalachian counterparts, which suggests that oral health literacy may follow the
same subpar pattern.
Kentucky continuously ranks among the highest nationally for edentate adults
amid reports of 13% of the adult population being completely toothless and increasing up
to 40% across some Appalachia areas for those of retirement age (Dawkins et al., 2013;
National Institute of Dental and Craniofacial Research, 2009). It is estimated that among
Kentucky children, almost 50% suffer from an average of two untreated dental caries,
with severity increased among children living in the most eastern areas, which are part of
the Appalachian region (Dawkins et al., 2013).
Gap in Knowledge
Although it is established that Kentucky Appalachians do suffer from a lower
level of health literacy (Ludke et al., 2006; Moser et al., 2015), there is little to no
documentation that investigates for a relationship between low oral health literacy and the
overall poor dental health status that this population is suffering from. There is also little
to no documentation investigating how self-oral hygiene techniques are used in the
Kentucky Appalachian population although research shows that self-oral hygiene
5
techniques such as teeth brushing are imperative to good oral health (Van der Weijden &
Slot, 2010). The study presented here will address this gap by evaluating for a
relationship between the oral health literacy level and how frequently oral hygiene
techniques are utilized in a Kentucky Appalachian population. Additionally, these
Kentucky Appalachian adults will be polled on their beliefs regarding the personal risk
poor oral health poses to them and their children, as well as their opinions and attitudes
towards oral hygiene techniques.
Need for Study
Historically accepted reasons for the poor oral health status in the Kentucky
Appalachian region have included a lack of dental insurance coverage as well as
decreased access to dental care professionals (Oral Health in Kentucky Technical Report,
2016). With the introduction of the Affordable Care Act as well as a Kentucky Medicaid
expansion program, over 100,000 more Kentuckians sought and received dental services
in 2014 than in 2013, indicating that with additional dental coverage now in effect, more
people are seeking to utilize these benefits (Oral Health in Kentucky Technical Report,
2016). However, this represents not even a fifth of the some 560,000 people who were
newly enrolled for Medicaid benefits during this time according to the Centers for
Medicare and Medicaid Services (as cited by Oral Health in Kentucky Technical Report,
2016). Certain areas of Kentucky have limited or no access to dental services as dental
professionals are more likely to cluster around more urban locations that have a higher
average socioeconomic status as these areas offer a potentially higher income (Saman,
Arevalo, & Johnson, 2011). This leaves the more rural areas in Appalachia underserved
6
or in some cases, with no dental professional service access as these areas have a lower
average socioeconomic status and so offer a potentially lower income for dental
professionals.
In 2010, the Kentucky Department for Public Health had issued recommendations
to improve dental professional coverage. These included improving recruitment of dental
students to areas of greater need and investigating the establishment of a dental school
located in the mountainous Appalachian region of Kentucky (Kentucky Department of
Public Health, 2010). To date, the ratio of dentists to patients in Kentucky rose from 5.6
per 10,000 population in 2006 to 6.0 dentists per 100,000 in 2015 (Kentucky Department
of Public Health, 2010; Oral Health in Kentucky Technical Report, 2016). However, the
pattern of unequal coverage still exists and some areas continue to go without any dental
professional access.
By investigating for a relationship between oral health literacy level and oral
hygiene habits, as well as evaluating for the attitudes towards oral health in the
Appalachian population, this could give public health officials insight into how Kentucky
Appalachians currently process and understand oral health information. The results may
suggest a need for interventions regarding proper oral hygiene habits as in some areas
these techniques may be the main source of currently available dental decay prevention.
Evaluating for the attitudes held by Appalachian caregiver’s towards dental disease and
oral hygiene habits in children will also be valuable information to obtain. This will give
insight into how parents and guardians are passing oral hygiene information and habits to
children as documentation shows that children largely learn health behaviors and
7
attitudes from their parents and guardians (Rhee, 2008). This is especially true of
preschool aged children as they depend solely on their caregiver for all health related
treatment and attention. By investigating the information regarding oral health that is
being passed from adult to child, opportunity may be found to improve the transition of
knowledge which may in turn help to break the cycle of poor oral health that is occurring
in the Kentucky Appalachian region.
Problem Statement
Despite increased dental insurance coverage as provided by a Kentucky Medicaid
expansion and an increase in the number of practicing dental health professionals, poor
dental health remains a severe public health threat especially in the more rural
Appalachian areas (Kentucky Department of Public Health, 2010; Oral Health in
Kentucky Technical Report, 2016). Dawkins et al. (2013) found that 49.7% of the school-
aged children included in their cross-sectional study suffered from an average of two
untreated dental caries. This is 16% higher than the estimated 33% national rate of
untreated dental caries in school-aged children, indicating that Kentuckians have an
increased rate of developing this health risk (Kandel, Richards, & Binkley, 2012). Having
poor childhood oral health increases the risk of having poor oral health as an adult, which
is a trend seen in Appalachia with some areas reporting upwards of 40% of their 65 and
older population being edentate (Dawkins et al., 2013; National Institute of Dental and
Craniofacial Research, 2009).
Current preventative efforts in decreasing the rate of dental decay occurrence in
Kentucky Appalachian children include the use of early grade school-based sealant
8
programs (Dawkins et al., 2013). To provide maximum benefit, sealants require routine
inspection and repair from an oral health professional. This leaves many children in
Appalachian areas where dental care access is low at increased risk of poor protection
(National Institute of Dental and Craniofacial Research, 2012; Reed, 2016). Sealants
wear away over time which may place children at increasing risk of dental decay onset as
they age (National Institute of Dental and Craniofacial Research, 2012). Sealants also do
not promote the use of proper oral hygiene habits, which are arguably the most important
aspects in dental decay prevention (Kidd, 2011). With no long-term benefits or education
regarding proper oral hygiene habits, the school-based sealant programs may only be
prolonging the onset of dental decay in Kentucky children. There is also evidence that
children are presenting to participate in sealant programs who already are suffering from
moderate to severe dental decay (Dawkins et al., 2013). This indicates that poor oral
health is occurring in preschool aged children, making it impossible for school-based
sealant programs to target and assist in preventing dental decay in these cases. Being too
young to do so for themselves, preschool aged children are dependent on their caregiver’s
for decisions regarding their oral health (Talekar, Rozier, Slade, & Ennett, 2005). Dye,
Vargas, Lee, Magder, and Tinanoff (2011) demonstrated that there is a strong
relationship between a mother’s oral health status and that of her dependent children. Dye
et al. (2011) found that mother’s with untreated caries were three times more likely to
have preschool aged children who also had caries when compared to children whose
mother’s did not have caries. This suggests that caretakers of preschool aged children
who have a higher level of oral health literacy are more likely to have children who do
9
not suffer from dental decay or untreated dental caries. With this documented link
between parent and child oral health status, it becomes clear that it is important to
evaluate for the beliefs, attitudes, and habits that parents hold towards oral health as they
may be passing negative habits and ideologies to their children. These aspects need to be
investigated and documented in Kentucky Appalachian parents to give public health
officials possible new venues of intervention. This is important as sealant programs,
increased state based dental insurance, and a rise in available dentists are thus far not
decreasing the rate of dental decay in this population.
Purpose of Study
The purpose of this study is to determine if there is a relationship between the
level of oral health literacy in Appalachian adults and the use of oral hygiene techniques
both in themselves and their dependent children. This will be done by comparing how
important Appalachian adults feel dental health is to how often they practice and enforce
oral hygiene techniques in the home. Sequentially, this study will also provide a
description of the current oral hygiene habits of Kentucky Appalachians of which there is
little to no current documentation available. This study will also explore the extent to
which Kentucky Appalachian adults feel they are in control of their own dental health, or
if they feel the only way to have proper dental health is solely by obtaining treatment by a
dental professional.
Independent Variable
The independent variable in this study is the level of oral health literacy expressed
by the Appalachian adult participants.
10
Dependent Variables
Two dependent variables in this study are the frequencies that caregiver’s use oral
health hygiene techniques personally and how often they enforce the use of these
techniques in their dependent children. Both frequencies will be evaluated with the
independent variable to identify any trends in the data. This evaluation is important as it
will show if caregivers with a higher level of expressed oral health literacy are utilizing
oral hygiene methods more often than the caregivers with a lower level of expressed oral
health literacy. The two groups of frequencies will then be compared to each other to
determine the existence of any differences in the rate of use of oral hygiene techniques
between caregiver’s and children. This comparison will serve to display if caregivers are
enforcing the use of oral health hygiene methods in their dependent children at higher
rates than they are personally using them. If no difference in rates is identified, this may
suggest that caregivers are passing their own oral health habits on to their children. This
would provide evidence that supports the theory that poor dental health is a cycle being
passed on from generation to generation.
Another dependent variable in this study is the extent to which the participants
feel that they are in control of their own dental health status and that of their dependent
children. The reported level of control will be compared to the independent variable of
expressed oral health literacy level in order to identify any trends or correlations between
the two variables.
The final dependent variable of this presented study is the perceived importance
of oral hygiene methods. The level of importance that each participant feels towards oral
11
hygiene methods both in themselves and in their children will be recorded and compared
with their expressed level of oral health literacy to determine if any trends exist in the
collected data. This information will also provide an overview as to how the included
Appalachian population currently feels towards oral hygiene methods.
Predictor Variable
This study will record if participants or their dependent children have ever had a
dental cavity. This predictor variable will be used in conjunction with other variables to
determine the independent variable of expressed oral health literacy, as well as to give a
description of the overall oral health status of the included population.
Research Questions and Descriptive Items
Research Questions:
1. Is there an association between the participant’s expressed level of oral health
literacy and the frequency at which they practice oral hygiene techniques?
Ha1: There is an association between the participant’s expressed level of oral
health literacy and the frequency that they practice oral hygiene techniques.
H01: There is no association between the participant’s expressed level of oral
health literacy and the frequency that they practice oral hygiene techniques.
2. Is there an association between the participant’s expressed level of oral health
literacy and the frequency at which their children practice oral hygiene techniques?
Ha2: There is a relationship between the participant’s expressed level of oral
health literacy and the frequency that their children practice oral hygiene
techniques.
12
H02:There is no relationship between the participant’s expressed level of oral
health literacy and the frequency that their children practice oral hygiene
techniques.
3. Are Kentucky Appalachian adults and children practicing oral hygiene techniques at
the same frequencies?
Ha3: There is a difference in how often oral hygiene techniques are used between
adults and children.
H03: There is no difference in how often oral hygiene techniques are used
between adults and children.
4. Were there differences in the survey responses gathered from the transitional
Kentucky Appalachian county and the distressed Kentucky Appalachian county?
Ha4: There are differences in the survey responses collected between the
transitional and distressed counties.
H04: There are no differences in the survey responses collected between the
transitional and distressed counties.
Descriptive Items:
1. Do Kentucky Appalachian adults practice oral hygiene techniques at frequencies
as recommended by the American Dental Association?
2. Do Kentucky Appalachian children practice oral hygiene techniques at
frequencies as recommended by the American Dental Association?
3. Do Appalachian adults perceive dental decay as a health risk?
4. Do Appalachian adults feel that dental decay is a preventable health risk?
13
5. Do Appalachian adults feel that they can personally decrease the risk of dental
decay?
6. Do Appalachian adults perceive oral hygiene techniques as important for good
health?
7. To what level do Appalachian adults perceive that enforcing childhood oral health
techniques will decrease the risk of dental decay as their children age?
8. Do Appalachian adults perceive poor dental health as a normal event?
The hypotheses and research questions will be answered by data collected from a
survey personally distributed in a sample Kentucky Appalachian population. Survey
items will be primarily presented using a Likert scale response format. An example of
this scale would be using 1 - 5, with 1 being ‘completely disagree’ and 5 being
‘completely agree’. The corresponding number of each rate will then be used to
statistically analyze the data set in order to determine answers for each hypothesis and
research question.
Theoretical Framework
The theoretical framework for this study is found in the health belief model
(HBM). The HBM was initially developed in the early 1950s by social psychologists in
order to explain why people fail to partake of programs that are designed to prevent and
detect disease (Glanz, Rimer, & Viswanath, 2008). Over time, the HBM was adjusted and
extended by others in order to examine the behaviors and attitudes people hold towards
health risks (Glanz et al., 2008). This theory takes into consideration the perceived
susceptibility and severity a person believes a health risk poses, as well as their perceived
14
benefits of a health habit and their perceived barriers to adopting the health habit. The
HBM also includes the construct of self-efficacy, or the personal belief that a person can
adopt an appropriate health habit in order to obtain the desired outcome, such as avoiding
a health risk (Glanz et al., 2008). These constructs are used to predict why people decide
to adopt or reject health habits that can assist in disease or health risk prevention. A
deeper analysis of the HBM can be found in Chapter 2.
The HBM theorizes that people with a higher level of perceived personal disease
risk along with a higher level of belief towards the seriousness of the disease, are more
likely to adopt a health behavior or habit that decreases the risk of that particular disease
(Glanz et al., 2008). These two constructs are included in this study by evaluating for
how caregiver’s view dental decay as a health risk in themselves and their dependent
children. The HBM predicts that if caregiver’s are found to believe that dental decay is
serious health risk, they will be more likely to adopt positive oral health habits to
decrease the risk of dental decay. However, if it is found that caregiver’s do not see
dental decay as a potential health risk, the HBM predicts that they will be less likely to
adopt positive oral health habits. If this is found to be the case in this study, this may
suggest that education is needed regarding poor oral health being a health risk to
encourage the use of proper oral hygiene techniques. The construct of perceived benefits
is also included in this study. This construct uses the belief regarding the positive aspects
of a health behavior to predict the behavior’s use. People who believe a health behavior
will provide a decreased risk of disease are more likely to adopt the behavior (Glanz et
al., 2008). By analyzing the beliefs held by Appalachian caregiver’s towards oral hygiene
15
techniques as dental decay prevention, it may be predicted whether they would be likely
to adopt more positive oral health behaviors.
Nature of the Study
This study will utilize data collection by way of a personally delivered survey
method design. Survey data collection has long been used as a way to gather data that is
representative of a population (Bartlett, Kotrlik, & Higgins, 2001). Surveys are a versatile
method of data collection as they can gather data that can be used to produce statistical
calculations and comparisons, but can also gather data that provides a descriptive
overview of an attitude, belief, health risk or issue (Fowler, 2014).
Definitions
Appalachia: Refers to the 205,000 square mile region that encompasses the
Appalachian Mountain range and includes portions of 13 states, from Southern New
York to Northern Mississippi (Appalachian Regional Commission, 2016). This region is
home to some 25 million people (Pollard & Jacobsen, 2011). The Kentucky Appalachian
region includes 54 of its easternmost counties, in which approximately1.2 million
Kentuckians live (Pollard & Jacobsen, 2011). Thirty-seven Kentucky Appalachian
counties are considered to be in a distressed state, with poverty rates that are up to three
times the national average (Commonwealth of Kentucky, 2015). Kentucky has
historically contained more distressed counties and communities than any other
Appalachian state and continues to display this unfortunate trend (Commonwealth of
Kentucky, 2015).
16
Health literacy: Describes a person’s ability to obtain, understand, and utilize
health related information and directions. Health literacy level can affect three different
key points of healthcare: access and use of health information, the patient-medical
professional relationship, and self-care (Paasche-Orlow & Wolf, 2007). A low level of
health literacy has been linked to a lower level of overall health as well as with an
increased rate of negative health outcomes (Nutbeam, 2008; Paasche-Orlow & Wolf,
2007).
Oral health literacy: Part of a person’s overall health literacy status, oral health
literacy refers to how a person absorbs and understands information pertaining to oral
health. Having good oral health literacy is believed to be a critical and necessary aspect
for people to have in order to improve their oral health (Horowitz & Kleinman, 2008).
Oral hygiene techniques: Includes brushing and flossing. When utilized correctly,
these methods arguably provide the most important and effective form of defense against
dental decay and disease (Kidd, 2011).
Assumptions
As data collection will take place within the Appalachian region, an assumption
of this study is that all included participants are residents of the Kentucky Appalachian
area. No data will be collected as to how long each participant has resided in Appalachia,
or if they have moved away from Appalachia and then returned during their lifetime. A
second assumption of this study is that the adults involved serve as primary caregiver’s,
have legal custody of, and live in the same household as their children. No data will be
collected that could serve as proof of these assumptions.
17
Scope and Delimitations
With Kentucky’s long history of poor oral health in its Appalachian communities
continuing into present day, this health risk is still a severe public health threat that keeps
encompassing generations. Some documentation suggests that targeting Kentucky
children, such as with dental sealant programs, is the best way to break the cycle of poor
dental health that is occurring (Reed, 2016). However, children who already have dental
decay are presenting to school-based sealant programs, suggesting that this process is
beginning in the home while the children are very young and dependent on their
caregiver’s for proper oral hygiene. Considering this, targeting young children
themselves may not be the most successful path in breaking the cycle of poor oral health.
In response to this realization, the study presented here is designed to evaluate the beliefs
and attitudes towards oral health that are held by Appalachian adults. This will provide
information into how Appalachian adults utilize oral hygiene techniques as it may be this
group that would provide the most benefit in breaking the cycle of poor dental health if
targeted with public health intervention efforts.
The boundaries of this study are firmly centered on the Kentucky Appalachian
population. Kentucky has a high rate of fluoridated water supplies, dental sealant
programs, and dental professional utilization (Saman et al., 2011). However, in
Kentucky’s rural Appalachian population there are areas where dental disease and
edentulous rates exceed the national averages, placing this population at an increased risk
for poor dental outcomes (Saman et al., 2011). The Kentucky Appalachian region is host
to some of the nation’s socioeconomically poorest communities who also suffer from
18
sporadic and uneven access to both dental insurance coverage and dental care services.
These factors have assisted in creating a large health disparity in the oral health status
between Kentucky Appalachians and their non-Appalachian counterparts.
The data from this study could potentially be used to address oral health
disparities that exist in Appalachian areas beyond the Kentucky region. The Appalachian
region across all 13 states have a common theme in that the people located here make less
than the national per capita income, and also experience higher rates of unemployment
particularly in the counties that are defined as being in a distressed state (Appalachian
Regional Commission, 2016). The Appalachian region also suffers from lower than
national levels of high school completion, indicating that a lack of education is present in
most Appalachian communities, again especially in the communities that are considered
distressed (Appalachian Regional Commission, 2016). These shared sociodemographic
trends suggest that the Appalachian population is similar regardless of state lines, making
it possible to make generalizations in one area that could apply to other Appalachian
areas, particularly in counties that share the same categorization such as distressed.
Limitations
A limitation of this study is that it requires the participants to have a basic level of
literacy in order to read and fill out the study survey. The requirement of self-reading is
necessary to avoid any bias towards answers that a second party may inflict with tone or
stance, be it unintentional or otherwise. To address this limitation, the survey will use
basic language to make it as easy as possible to understand what is being asked. The
19
directions will also be written in a clear manner and face-to-face explanation of the
directions will also be available to each participant if needed.
Another limitation is that data is being collected from only two Kentucky
Appalachian communities. Having data from all 54 counties that make up the Kentucky
Appalachian region would supply much more in-depth data regarding the oral health
habits and beliefs that are present. However, that is beyond the scope of this study. This
limitation will be acknowledged and addressed by assuming the generalizability of results
only to other Appalachian communities that hold similar sociodemographic traits to the
two that are included in this study.
Significance
The primary goal of public health officials is to prevent a disease or health event
from occurring, as opposed to the medical field which is largely more focused on
treatment after the health event has occurred. To do this, public health officials must
understand the factors which may be influencing or leading to the health event occurring
(American College Health Association, 2016; Centers for Disease Control, 2015) In
understanding these factors it is useful to utilize the social-ecological model which takes
into consideration the relationships between individual, community, and societal factors
which may be influencing the rate of a health event in any population (Centers for
Disease Control, 2015). This proposed study is focused on providing individual level data
to help further understand the factors that may be influencing the rate of dental decay and
disease that is occurring in the Kentucky Appalachian population. The collected data may
be valuable in assisting to bring social change to the community as only by understanding
20
currently held attitudes and beliefs towards a health event can we hope to modify and
transform those attitudes and beliefs into more positive health behaviors over time
(Leicht, 2013). To achieve this, the proposed study will supply data that could be of use
to public health officials when designing and implementing oral health programs and
interventions in the Kentucky Appalachian region. If a relationship can be identified
between the level of oral health literacy of adults and the usage rate of oral hygiene
techniques being enforced in their dependent children, this may suggest that caregiver
education is needed to increase the usage rate of oral hygiene techniques A higher usage
rate, especially in the very young, may decrease the number of children that are
presenting to school-based sealant programs with dental decay already present. Raising
awareness about the benefit of proper oral hygiene techniques and the utilization of such
practices may help decrease the rate of poor oral health, particularly in communities
where dental health professional access is low to nonexistent. By targeting the oral health
literacy level of Appalachian adults, this may assist public health officials in breaking the
cycle of poor oral health that is so prevalent in this population.
Summary
The Kentucky Appalachian population has long suffered from an overall poor
level of oral health. While current efforts at eliminating this health disparity have
included increasing dental insurance access and targeting school-aged children with
sealant programs, this study suggests that a more effective target for intervention would
be found in improving the oral health literacy level of Appalachian adults. In support of
this suggestion, data will be collected that will provide an overview of the current usage
21
rate of basic oral hygiene techniques and evaluate for a relationship between this usage
rate and the expressed level of oral health literacy of Appalachian adults. The attitudes
and beliefs held by the Appalachian participants towards oral health will also be
documented to determine if education could be used to improve currently held ideals.
This data could be valuable in creating public health interventions to assist in breaking
the cycle of poor oral health that is present in the Appalachian community, thereby
helping to introduce a positive social change to this population.
22
Chapter 2: Literature Review
Introduction
The poor oral health status of the Appalachian region is a much documented and
established disparity (Polk et al., 2008). Even with advancements in water fluoridation
efforts, school-based sealant programs, and an increase in dental insurance coverage,
Kentucky repeatedly places among the nation’s highest for untreated childhood dental
decay and edentate adults (Dawkins et al., 2013; Kandel et al., 2012; Oral Health in
Kentucky Technical Report, 2016). This information suggests that there are additional
factors in existence that are contributing to the poor oral health status of the Kentucky
Appalachian population. This study seeks to investigate for additional factors by
evaluating for a relationship between the level of oral health literacy of Kentucky
Appalachian adult’s and their use of oral hygiene methods both in themselves and their
dependent children. Additionally, the perceived level of benefit that these caretakers feel
that these methods provide will be recorded.
In this chapter, information is presented that shows how this study is largely
driven by the key concepts of the Health Belief Model. It also presents detailed data
gathered from past and current research that supports the main focus and ideals of this
study, as well as describes gaps that have been identified in the currently available
literature. In cases where little established information is available, argument is made for
the inclusion of, evaluating for, and providing information on such under-studied
concepts.
23
Literature Search Strategy
In the development of this proposal, several different online databases and search
engines were used to gather supporting information:
Table 1
Databases and search engines used in this study.
Appalachian Regional Commission
Behavioral Risk Factor Surveillance System 2012
Centers for Disease Control and Prevention
Google Scholar
Phys.Org
PubMed: U.S. National Library of Medicine, National Institutes of Health
United States Census Bureau 2010
University of Kentucky Online Library
Walden University Online Library
In most cases, a search engine was used in conjunction with a database. For instance, an
abstract may have been found on PubMed but due to limited student usage rights, the
actual research paper was then pulled from either the University of Kentucky or Walden
University online library database so that the paper could then be perused in its entirety.
Many search terms were used to find appropriate and available supporting
documentation within the included databases and search engines. When applicable, the
same search terms were applied to each of the included data bases and search engines to
ensure an in-depth search of available sources. The following is a non-inclusive overview
of the major key search terms and phrases used in data discovery:
Appalachian health literacy
Appalachian dental health status
Appalachian oral health status
24
Appalachian region demographics
Construct of perceived barriers
Construct of seriousness
Construct of susceptibility
Constructs of the health belief model (HBM)
Dental hygiene
HBM
HBM development
Kentucky Appalachian health literacy
Kentucky Appalachian region demographics
Kentucky dental health status
Kentucky economic status by county
Kentucky health literacy
Kentucky population by county
Kentucky oral health literacy
Kentucky oral health status
Oral hygiene
Oral hygiene habits in Appalachia
Origin of the health belief model
Perceived dental health
Perceived oral health
Self-efficacy
25
Self-efficacy in Appalachia
All searches pertaining to oral health or ideals were performed twice: once with using the
term ‘oral’ and once using the term ‘dental’. This was to minimize the risk that sources
may be missed due to the use of differing professional terminology.
Where appropriate and available, literature was included that was published in the
standard most recent 10-year timeframe. However, there were instances where older
information had to be included. For example, when describing the Health Belief Model,
current sources of information were unclear as to the exact origins and development of
this theory. Older, original work had to be located in order to give a clear understanding
of the background to this theory and how it drives this study. In some cases, such as
establishing the current picture of the health literacy status of Kentucky Appalachians,
data that is right on the 10-year inclusion cut-off mark were included as no updated
information could be found. To my knowledge, these are the most recent sources
documenting this concept in the area of interest. Additional, more recent resources were
included to support the overall status, but these resources do not use Kentucky
Appalachia as their focus but instead other areas of Appalachia. In one case, no specific
documentation could be found to use to support or describe the included concept. For this
instance, argument was made on the importance of this concept and linked it as a possible
factor behind the results of another study, although this study did not specifically mention
the concept.
Much effort was made to include supporting literature that was only from peer-
reviewed sources. In instances were such literature was not appropriate, such as
26
determining population size or the economic status of Kentucky Appalachian areas,
sources were used from federally and state funded or controlled sites. Examples of these
sources are the data included that originates from the Appalachian Regional Commission,
the United States Census Bureau, or the Fluoride Action Network.
Theoretical Foundation
The established theoretical framework that is being used to support this study is
the Health Belief Model (HBM). In the 1950s, the United States Public Health Service
(PHS) existed to prevent disease and health risk as opposed to providing any treatment
for established diseases (Rosenstock, 1974). Thus, the Public Health Service did not yet
take into consideration any issues that were caused by a person’s compliance with
medical directions, health literacy level, or a lack of communication between medical
professionals and patients: the PHS only focused on preventative efforts (Rosenstock,
1974). During this time, it was clear that public health prevention programs were being
met with limited participation and success (Strecher & Rosenstock, 1997). This trend was
particularly evident in the low participation rates of the then available tuberculosis (TB)
and dental disease screening tests, with continuation in the later introduced rheumatic
fever, polio, and influenza screening and prevention efforts (Rosenstock, 1974). In 1958,
Hochbaum presented probability samples taken of adults living in cities that had
conducted free TB screening programs in mobile X-ray centers (as cited by Strecher &
Rosenstock, 1997). In this report, Hochbaum included the belief that participants had
towards how susceptible they were to contracting TB as well as their belief towards
perceived personal benefit of early TB detection and diagnosis (as cited by Strecher &
27
Rosenstock, 1997). Hochbaum found that participants who displayed beliefs both in
perceived personal susceptibility and benefit of early detection were four times more
likely to have had a voluntary chest X-ray screening as those who displayed no beliefs in
either category (as cited by Strecher & Rosenstock, 1997). Strecher & Rosenstock (1997)
contend that this work by Hochbaum laid the ground work for the HBM in contributing
the first two included constructs of perceived susceptibility and perceived benefits, and
showing how these constructs can be used to determine how likely people are to partake
of public health efforts. The work by Hochbaum was of considerable contribution in that
it took into consideration the personal beliefs people hold towards a health risk, a concept
that until then had been overlooked by the PHS.
Over the next several decades, further investigations by many different
researchers helped to expand and clarify the two constructs identified by Hochbaum into
the HBM that is used today (as cited by Strecher & Rosenstock, 1997). The HBM is now
recognized to utilize four main constructs which include the perceptions of susceptibility,
severity, benefits, and barriers (Glanz et al., 2008; Hayden, 2014). Any of these
constructs can be used alone or in combination to explain a person’s health behavior or
habits (Hayden, 2014). The HBM also includes the concepts of cues to action and self-
efficacy that serve as additional factors to the four main constructs (Hayden, 2014). The
HBM constructs can also be influenced by modifying variables such as cultural habits
and beliefs, education level, past experiences, age, motivation and other such personal
demographics (Hayden, 2014).
28
The four main constructs that make up the HBM all deal with different perceived
beliefs that a person may hold towards a disease or health risk. The first construct of the
HBM is perceived susceptibility which is described as the greater transitional a person
feels from a disease or health event, the more likely they are to adopt health behaviors or
habits that may decrease that risk (Hayden, 2014; Rosenstock, Strecher, & Becker, 1988;
Strecher & Rosenstock, 1997). Although this seems a very strong indicator construct in
terms of assessing how likely people are to adopt a health habit or behavior, Carpenter
(2010) discovered differently. Through a meta-analysis of published studies utilizing the
HBM, Carpenter (2010) found that susceptibility alone to be the weakest predictor for a
person’s behavior. This finding contradicted the earlier established belief that
susceptibility was a strong indicator and predictor for health prevention behavior and
treatment (as cited by Carpenter, 2010; Janz & Becker, 1984).
The construct of severity serves to describe the level of seriousness a health risk is
perceived as posing (Strecher & Rosenstock, 1997). This construct theorizes that the
more serious risk a person perceives a disease or health event as posing, the more likely
they are to adopt preventative health behaviors. While this construct can be influenced by
medical knowledge, it is possible that the perceived seriousness towards a health risk is
also based on a person’s accepted cultural norms or personal experience (Hayden, 2014).
Carpenter (2010) found that in general severity alone was a poor indicator of whether a
person would adopt a health behavior or habit, which was similar to previous findings by
Harrison, Mullen, and Green (1992). However, when only studies that included taking
prescription drugs were included in the analysis, it was found that there was a strong
29
relationship between severity and whether the drug regime was adopted (Carpenter,
2010). This suggests that this construct can be successfully used under certain
circumstances, such as with predicting the adoption of drug therapy.
The construct of benefits describes a person’s opinion of how well a new health
behavior or habit will work in decreasing a health risk (Hayden, 2014; Joseph, Burke,
Tuason, Barker, & Pasick, 2009). Carpenter (2010) found that the construct of benefits
was a more effective predictor for prevention behavior than for treatment behavior. This
suggests that prevention behaviors are more likely to be implemented before the
occurrence of a health event than treatment behaviors that are recommended after a
health event occurrence. The construct of benefits can also include the influence of
perceptions that are not specifically health related. For instance, the financial gain or cost
that a health behavior may entail could also be influential on how likely a person is to
adopt the behavior (Glanz et al., 2008).
The last construct of the HBM is that of perceived barriers. This construct
includes the perceived potentially negative aspects a person may use as reasons not to
adopt a particular health behavior (Glanz et al., 2008). Barriers may include monetary
expense, side effects, inconvenience, or time. Carpenter (2010) found that perceived
barriers alone was the most influential of the four constructs when predicting the
likelihood of health behavior adoption.
The concept of cue to action was included in the development of the HBM as an
additional variable that could influence the other constructs and therefore also influence
the likelihood of health behavior adoption (as reported by Carpenter, 2010; Glanz et al.,
30
2008). A cue to action can be from an external environmental influence such as viewing a
media campaign or reading a pamphlet, or an internal influence such as a sneeze
prompting the use of sinus medication (Carpenter, 2010; Glanz et al., 2008). The concept
of self-efficacy was not included in the early formulations of the HBM but has since
become an accepted variable to consider when predicting the adoption of health
behaviors (Glanz et al., 2008). Self-efficacy is the confidence a person has to how
successful they will be in conducting a health behavior well enough to obtain the desired
outcome (Glanz et al., 2008; Schunk & Pajares, 2009). Since its inclusion, self-efficacy
has been recognized as factor that differs among developmental levels and cultures and
that can influence a person’s education and career decisions as well as their health related
decisions (Schunk & Pajares, 2009).
Used alone, Carpenter (2010) found the constructs of seriousness and
susceptibility to be the poorer behavior adoption predictors of the HBM, which
contradicted earlier conducted findings. However, he contends that these findings could
be in error due to the following factors:
1. Susceptibility was difficult to measure in level as those who have experienced a
health event do not vary in their perception of susceptibility: they are susceptible.
Including those who were already diagnosed with a disease or health event likely
skewed his results.
2. Self-efficacy is argued to influence and moderate the constructs of seriousness
and susceptibility. Carpenter (2010) did not allow for this influence in his meta-
analysis.
31
3. Carpenter (2010) stipulates that seriousness and susceptibility may be moderated
by each other which is not considered in his study.
Therefore, the constructs of seriousness and susceptibility along with benefits are
included in the presented study and will be used together as each holds applications
which are the focus of this study. The construct of perceived susceptibility can be used to
define populations who may be more at risk of a health event (Glanz et al., 2008).
Primarily, this study will evaluate for how Appalachian adults view dental decay as a
health risk in themselves and in their dependent children, thereby identifying a possible
population who may be more at risk of dental decay due to a low perceived susceptibility.
Descriptive items 3, 4, and 5 are specifically related to measuring the level of
susceptibility this population may feel they are at from dental decay:
- Do Appalachian adults perceive dental decay as a health risk?
- Do Appalachian adults feel that dental decay is a preventable health risk?
- Do Appalachian adults feel that they can personally decrease the risk of dental
decay?
By answering these questions, an overview of the population’s perceived risk of dental
decay may be established.
The construct of seriousness is included in this study as it is a vital concept on its
own, but may also contribute to the population’s perceived susceptibility to dental decay.
Descriptive items 6 and 8 investigate how important the population views dental health as
a part of their overall health level:
32
- Do Appalachian adults perceive oral hygiene techniques as important for good
health?
- Do Appalachian adults perceive poor dental health as a normal event?
Gathering this data will provide an overview into how serious the included population
views the risk of poor dental health. The perceived seriousness may also tie into the
perceived susceptibility to dental decay that this population may demonstrate. For
instance, if it is found that the included population indicates that poor dental health is a
normal event, this may contribute to a higher level of perceived susceptibility: the event
is normal, it will likely occur, and so the population feels more at risk. In this way, these
two constructs will work together and overlap in the research questions to give a more
detailed and effective overview of the included population.
The construct of perceived benefits will also guide this study. An application of
this construct is defining the actions that may bring about the desired health outcome
(Glanz et al., 2008). In this study, the desired outcome is improved dental health.
Research questions 1, 2, and 7 relate to measuring the level of benefit that Appalachian
caregiver’s feel oral hygiene habits will contribute to their overall dental health level:
- Do Kentucky Appalachian adults practice oral hygiene techniques at frequencies
as recommended by the American Dental Association?
- Do Kentucky Appalachian children practice oral hygiene techniques at
frequencies as recommended by the American Dental Association?
- To what level do Appalachian adults perceive that enforcing childhood oral health
techniques will decrease the risk of dental decay as their children age?
33
A lower rate of use of oral hygiene habits may indicate a lower perceived benefit level in
the included population.
By using these constructs together, the fundamental foundation of the HBM may
be applied to this study to predict the needs of this population. Using the HBM, it may be
predicted that if caregiver’s are found to believe that dental decay is a serious health risk,
they may be more likely to adopt positive oral health habits in order to decrease their risk
of dental decay. These results may indicate a need of community education regarding the
adoption of proper oral hygiene habits. Conversely, if it is found that caregiver’s do not
see dental decay as a potential health risk for themselves or their dependent children, the
HBM would predict that this population is less likely to adopt positive oral hygiene
habits. These results would show a need for education regarding the risk of poor oral
health in order to encourage the adoption of proper oral hygiene techniques in this
population. Using the HBM can help predict the needs of a population, in this example by
predicting the focus of community education programs.
Literature Review Related to Key Variables and/or Concepts
Kentucky Appalachian Region
The Appalachian region is an area containing 205,000 square miles that follows
the Appalachian Mountains reaching across 13 states from Mississippi to New York
(Appalachian Regional Commission, 2016). Based on its national ranking position, each
of the 420 counties that are included in the Appalachian region is classified into one of
five economic status ranks (Appalachian Regional Commission, 2016). These rankings
are as follows:
34
Distressed: The worst 10 percent of the nation’s counties, these counties are the
most economically distressed areas.
At-risk: Ranked between the worst 10 to 25 percent of the nation’s counties and
are at risk of becoming distressed.
Transitional: These counties are seen as transitioning between strong and weak
economies and are made up of the worst 25 to the best 25 percent of the nation’s
counties.
Competitive: Competitive counties are those that are ranked in the best 25 percent
to 10 percent of the nation’s counties.
Attainment: The strongest of economies, these counties are those that rank in the
best 10 percent of the nation’s counties.
With 38 of its 54 Appalachian counties being ranked as distressed, Kentucky has
the largest number of distressed counties out of all the 13 states included in the
Appalachian region (Appalachian Regional Commission, 2015; Appalachian Regional
Commission, 2016). The highest ranking found in Kentucky Appalachia is shared by four
counties that have attained transitional status. However, all four of these counties hold
areas that are considered distressed, indicating that pockets of economically depressed
people are still present in these counties.
Much like other areas of Appalachia, the Kentucky Appalachian region suffers
from rates of dental disease and decay that exceed the national average, both in children
and adults (Dawkins et al., 2013; Kendal et al., 2012; Oral Health in Kentucky Technical
Report, 2016). To assist in combating poor dental health, Kentucky passed regulation in
35
1994 establishing mandatory water fluoridation for all water systems that are serving a
population of 3,000 or more (Fluoride Action Network, 2016). Legislature mandates that
communities between 1,500 and 3,000 are required to add fluoride to their water supplies
but only if the appropriate equipment is available from the Cabinet for Human Resources
(Fluoride Legislative User Information Database, 2012). From the 2010 United States
Census, it is estimated that there are 310 Kentucky towns containing 3,000 and under
inhabitants and while it is unclear if any of these are adhering to fluoridation guidelines,
it is reported that over 99% of Kentuckians have had access to fluoridated water systems
since 2006 (Fluoride Action Network, 2016).
School-based sealant programs are another effort made in the state of Kentucky in
order to decrease the rates of childhood dental decay. These can include actual plastic
sealants that are coated on the back teeth of children that act as a barrier to bacteria and
food particles, or fluoride varnishes, which are painted on the teeth to assist in hardening
the existing enamel (Madison County Health Department, 2016; Northern Kentucky
Health Department, 2016). School-based sealant programs were added to Kentucky
Medicaid as a preventative program in the 1990’s and target children when their first and
second permanent molars appear, ages 6-7 and 11-13 (Reed, 2016). To date, there are 23
local health departments who are participating in and practicing sealant programs in
Kentucky schools where 50% or more students are eligible for free or reduced cost
lunches, which are indicative of areas of the most need of health services (as reported by
Reed, 2016). In addition, the University of Kentucky Dentistry department assists in over
40 Kentucky counties by providing mobile units to various schools in order to provide
36
sealants to children of appropriate age (University of Kentucky Dentistry, 2012). It has
long been believed that sealants were the most effective way of preventing childhood
dental decay and disease (Dawkins et al., 2013; Reed 2016). However, new evidence is
emerging to refute this accepted fact. In the COHRA1 cohort study conducted by the
University of Pittsburgh using participants from Appalachian West Virginia, high rates of
dental decay were found in children even with the increased use of dental sealants
(University of Pittsburgh, 2016). The University of Pittsburgh also found that in some
areas, dental decay occurrence was happening in young children at 144% the rate
reported by the Center for Disease Control’s 1999-2004 National Health and Nutrition
Examination Survey (University of Pittsburg, 2016). The data provided by the University
of Pittsburg shows that dental sealants may not be decreasing the rate of childhood dental
decay, and that the rate of dental decay may be occurring in larger rates than previously
documented in some Appalachian areas, regardless of sealant use rates.
In a report conducted by Delta Dental of Kentucky and Kentucky Youth
Advocates, it was found that the oral health status of Kentucky school children is
worsening, even though access to oral health care has greatly improved over the last 15
years (as reported by Patrick & Thomas, 2016). This report found that the number of 3rd
and 6th
graders in need of early or urgent dental care rose from 32% in 2001 to 49% in
2016. They also acknowledge that children who reside in Appalachian areas
demonstrated the greatest need for urgent dental care. The report also established that
although sealant use in 3rd
and 6th
graders has increased by 14% since 2001, 50% of the
children included in the report were found to have had no sealants on any of their
37
permanent molars. This report and the findings by the University of Pittsburg suggest that
despite efforts to combat childhood dental decay have been in place, the rates of
childhood dental decay are increasing especially in populations of lower socioeconomic
status, such as Appalachian communities. These findings support the need for more
research in local communities to determine if any other opportunities exist for public
health officials to use in order to help in decreasing the rates of dental decay.
Independent Variable: Oral Health Literacy
The term oral health literacy was first defined and documented in the Healthy
People 2010 goals (as reported by Horowitz & Kleinman, 2008). It mirrors the concept of
health literacy in that it describes the level to which people can obtain, process, and
understand the basic information that is needed to make appropriate dental health
decisions (Horowitz & Kleinman, 2008; Jones, Lee, & Rozier, 2007). The Healthy
People 2010 report suggests that poor oral health literacy may be acting as a barrier to
proper dental health and is assisting in creating dental health disparities and poor oral
health outcomes (Centers for Disease Control and Prevention, 2011; Jones et al., 2007).
Milfrom, Garcia, Ismail, Katz, and Weintraub (2004) suggest that poor oral health
literacy is a national public health issue which worsens in areas of lower socioeconomic
and demographic status. People with low oral health literacy levels are less likely to
utilize preventative habits or dental care professionals, thereby contributing to higher
rates of dental disease and decay (Horowitz & Kleinman, 2008).
To date, it is unclear how oral health literacy is affecting the Kentucky
Appalachian population; however it is documented that these areas do suffer from low
38
health literacy levels (Ludke et al, 2006; Moser et al, 2015). Polk et al. (2008) document
that poor oral health is a shared trait over much of the Appalachian region. In their cross-
sectional study using Appalachian parent-child pairs in West Virginia and Pennsylvania,
they plan to investigate how individual, family, and community factors were contributing
to dental disease (Polk et al., 2008). Polk et al. (2008) wish to examine for a relationship
between oral hygiene habits such as brushing and flossing and rate of dental caries. This
study is still ongoing and so no final data are available, however Polk et al. (2008)
document the need for studies that investigate for factors that are contributing to the poor
oral health status of Appalachian communities as economic disadvantages alone have
been shown to have little overall impact on the rate of dental service use in this
population.
Guo et al. (2014) further document that people who do hold dental health
insurance are not utilizing preventative dental care services, particularly in more rural
areas and so economic status is not necessarily a sole barrier to proper dental health. In
this telephone survey based study that took place in rural Florida areas, Guo et al. (2014)
found that the influence of oral health literacy was just as an important factor on self-
reported oral health status as the standardized effects of gender, race, education, financial
status and the quality of patient to dentist communication. These findings show that oral
health literacy can be a factor in oral health status and may be influencing other perceived
barriers to proper oral health. A limitation with this study is that it was conducted in a
non-Appalachian population; however it supports the possibility that oral health literacy
may be affecting Appalachia in a similar fashion.
39
Miller, Lee, DeWalt, and Vann (2010) conducted a cross-sectional study of
young children and their caregiver’s who presented for care at the University of North
Carolina at Chapel Hill School of Dentistry. Here they collected data on the caregiver’s
oral health knowledge level, oral health behaviors, and the reported and clinical oral
health status of each child (Miller et al., 2010). The data showed that caregiver literacy
was significantly associated with their dependent child’s dental disease status.
Caregiver’s who demonstrated a lower level of oral health knowledge and behaviors were
more likely to have children who presented with dental disease (Miller et al., 2010).
Although the study does not specify designation, North Carolina does hold Appalachian
counties so it is possible that some participants shared the same overall demographics as
Kentucky Appalachian’s. The findings by Miller et al. (2010) suggest that a low level of
oral health knowledge, i.e. oral health literacy, may also be contributing to the very high
rates of childhood dental decay that is occurring in Kentucky Appalachian areas and that
an investigational study into this population may show similar results. Lee, Divaris,
Baker, Rozier, and Vann (2012) also conducted a study that associated oral health
literacy with oral health status in a North Carolina population. Their results mirrored that
of Miller et al. (2010) in that a lower oral health literacy level was associated with a
poorer level of oral health. Lee et al. (2012) contend that while much work exists to
associate health literacy with overall health status, efforts towards linking literacy to
dental health is a relatively new phenomenon. This suggests that more research is needed
to investigate oral health literacy in differing populations as there are gaps in the current
knowledge.
40
Dependent Variable: Oral Hygiene Methods
Oral hygiene methods include techniques such as brushing and flossing. When
used correctly, these methods arguably provide the most important and effective form of
defense against dental decay and disease (Kidd, 2011). The American Dental Association
(2016) maintain that brushing twice a day and flossing once per day are the most
effective oral hygiene methods people can utilize in preventing dental decay, both in
adults and children. Frisbee, Chambers, Frisbee, Goodwill, and Crout (2010) conducted a
cross-sectional convenience study that investigated for associations between oral hygiene
habits, obesity, and systemic inflammation in children from Appalachian West Virginia
communities. They collected the data from health screenings conducted at community
based facilities. Frisbee et al. (2010) conclude that preventive oral care in children is
important as oral health status is associated with other diseases. The work conducted by
Frisbee et al. supports the fact that proper oral hygiene methods are important to overall
health and that these methods should be investigated to establish if they are being used
correctly.
At the time of this study, Neiswanger et al. (2015) were conducting a longitudinal
study utilizing the Center for Oral Health Research in Appalachia (COHRA), which is a
collaboration effort between the University of Pittsburg and West Virginia University to
investigate the high rates of dental disease in these areas. Neiswanger et al. (2015) are
evaluateding for factors influencing the oral health of pregnant women and their babies
located in Appalachian areas of West Virginia and Pennsylvania and to date have reached
70% of their recruitment goal. The women recruited from West Virginia represent rural
41
communities versus the women form Pennsylvania which are considered urban. Thus far
the data collected by Neiswanger et al. (2015) are showing that women in West Virginia
are brushing their teeth at similar rates as women in Pennsylvania, but do not floss or see
a dentist at the same rate, have increased rates of dental disease, and have less education
and more unemployment. A major focus of this study is to investigate and follow the
dental health status of the children born during the study period to determine if the
children’s dental health status differs between the two groups of women, possibly
indicating that poor dental health may be occurring in children of very young age and that
this is setting a pattern that they will continue to follow throughout their life cycle.
The studies conducted by Kidd (2011) and Frisbee et al. (2010), as well as the
emerging data from Neiswanger et al. (2015) all support the importance of evaluating for
the use of oral hygiene methods across populations. If a deficit in use is found, this may
point towards an opportunity for community education promoting the use of proper oral
hygiene as a means in decreasing the rate of poor oral health that is occurring.
Dependent Variable: Perceived Control of Oral Health
The perceived control of oral health is the extent to which people feel that they
are in control of their oral health status. There is a distinct lack of current documentation
specifically on this concept. This study will evaluate for how people in an Appalachian
Kentucky community feel towards controlling their own oral health. While it is important
to obtain professional dental services for optimal oral health, it is equally as important to
perform oral hygiene methods in the home, such as brushing and flossing to properly
prevent dental disease. The study presented here will evaluate for how well people feel
42
they are in control of their oral health versus feeling that only by seeing a dental health
professional will they have proper oral health, meaning they believe only a dentist has the
ultimate control over their oral health status. It may be found that poor self-efficacy may
exist in this community: in this context, a lack of confidence in how successful people
feel they can conduct oral hygiene methods well enough to obtain proper oral health. It
has been documented that self-efficacy is a functional predictor to adopting and
maintaining a health behavior: the more confidence people have in conducting a health
behavior in such a way as to obtain the desired affects, the more likely they are to adopt
the health behavior (Schwarzer et al., 2007). Schwarzer et al. (2007) found that self-
efficacy was a better predictor of health habit adoption than health risk perception, or the
HBM construct of susceptibility. If it is found that low self-efficacy towards oral hygiene
does exist, this could provide valuable information when designing educational programs
and messages regarding Appalachian oral health. This may also tie into the findings of
Savage et al. (2014) in that some of their participants felt it was easier to simply ‘give in’
to poor oral health, possibly suggesting that those participants do not perceive that they
have control over their own oral health status.
Dependent Variable: Perceived Importance of Oral Hygiene
The perceived importance of oral hygiene is how important people think oral
hygiene habits are in promoting both good oral health, but also for their overall health
status. The perceived importance of oral hygiene is vital for understanding how people
think about and view oral hygiene habits and how much importance they place on such
43
habits. The more importance they place on these habits, the more likely they may be to
utilizing such habits in their daily routines.
A study was recently conducted with students attending a state university located
in an Appalachian Kentucky community (Savage, Scott, Aalboe, Stein, & Mullins, 2014).
The study consisted of 67 students participating in face-to-face focus groups and 587
students taking a survey. The results of this study lend support to the trend of poor dental
health that is found in Kentucky: 50.3% reported brushing twice per day, 17.6% reported
flossing once per day, and 23.9% reported that they had visible, active decay in their teeth
(Savage et al., 2014). In the focus groups, it was found that several of the participants felt
that poor dental health in Kentucky was not an accurate depiction of the true oral health
status and that this misconception was due to media portrayals (Savage et al., 2014).
However, some of these same participants went on to describe how many people they
knew from their home towns routinely never saw a dentist and they admitted that they
thought ‘some people’ didn’t understand the necessity of proper oral care (Savage et al.,
2014). Most participants stated the reason that they did not floss was that it is a time
consuming process, although some recognized that flossing is one of the best decay
prevention methods. Many participants described how it was easier to give in to the poor
oral health status of their communities and simply not care or place importance on proper
oral health (Savage et al., 2014). Wondering if good oral health was worth the effort it
requires was another point brought up by some participants (Savage et al., 2014). The
work by Savage et al. (2014) is an excellent source for how oral hygiene is perceived in a
rural area and results such as these can provide invaluable information when developing
44
programs targeting such populations. The limitation to this study was that it was
conducted at a state university and so both in state and out of state students participated:
there is no way to ensure that only students from Kentucky Appalachian communities
were included. While health disparities in Appalachia have been well documented,
Savage et al. (2014) maintain that research such as theirs that investigate the attitudes
towards such health disparities is what will provide invaluable information for message
design in future Appalachian health interventions.
Predictor Variable: Dental Decay
In this study, information will be obtained regarding if the included participants
have ever had dental decay in the form of dental cavities. The purpose of this data is two-
fold. First, it will serve to demonstrate if the included participants are typical of the
overall Appalachian region. It is documented that the Kentucky Appalachian population
has increased rates of dental cavities in both adults and children, and increased rates of
childhood decay (Dawkins et al, 2013; Oral Health in Kentucky Technical Report, 2016).
While it is beyond the scope of this study to physically examine participants for their oral
health status, self-reported data will be gathered to obtain a current overview of the
included participant’s oral health status. Secondly, this data will be used as a partial
indicator of the current oral health literacy level of the included participants and used as a
variable against other information, such as oral hygiene use frequencies.
Summary and Conclusion
After an extensive search of the available literature, many trends have been
identified and documented.
45
1. Although the overall Appalachian dental health status has much documentation
showing that it is a health disparity, little research can be found that exists to
explain this phenomenon.
2. Currently accepted reasons behind this disparity include lack of insurance
coverage and a lack of available dental health professionals, however new
research is emerging that shows that these reasons are not enough to explain why
dental health statuses continue to be poor, and in some cases worsen, in
Appalachian areas.
3. In the state of Kentucky, sealant programs are widely used in an effort to decrease
the rate of childhood dental decay. However, new research centering on a West
Virginia community suggests that sealants are not decreasing this health risk.
4. Although the concept of health literacy is widely documented as being directly
linked to a person’s health status, little research has been conducted that
investigates the concept of oral health literacy, especially in Appalachian areas.
5. Little is known about the oral hygiene habits utilized in Appalachia. Two large
cohort studies that included investigation for these habits were located, but they
are still in progress and so little data is available. These two studies do not focus
Kentucky Appalachia specifically, but areas of West Virginia and Pennsylvania.
Despite school-based sealant programs and widespread fluoridation efforts,
Kentucky continuously ranks among the nation’s highest for childhood dental decay and
edentate adults. A common factor among these efforts is that they supply little in the way
of education regarding proper oral hygiene habits or improving oral health literacy. These
46
factors combined with the gaps in current literature regarding how Kentucky
Appalachians view oral hygiene and oral health all point to a need for further research in
this population. If it can be established that there is poor self-efficacy regarding oral
health in Kentucky Appalachian communities, this may give public health officials
valuable knowledge in how to develop new messages and community education efforts
that target the dental health disparity occurring in Appalachia.
47
Chapter 3: Research Method
Introduction
The purpose of this research study was to assess for perceived attitudes towards
oral health as well as the usage rates of oral hygiene habits in two Kentucky Appalachian
communities. These data were sought as it is suspected that a lower level of oral health
literacy may be contributing to the elevated rates of poor oral health that exists in this
overall population.
In this chapter, the target population and the research design are thoroughly
presented. The methodology and the statistical analysis methods that will be conducted
on the gathered data are also presented and solidified. The projected sample size is
determined by taking into multiple factors, and the threats to the validity of the study are
examined and discussed.
Research Design and Rationale
The study presented here will follow a quantitative research design in that it will
be primarily developed to collect numerical data, which is the distinctive hallmark of the
quantitative research method (Creswell, 2014). The quantitative research method can be
further broken down into four main research types: experimental, quasi-experimental,
correlational, and descriptive (Center for Innovation in Research and Teaching, 2017;
University of Wisconsin, 2017). The experimental and quasi-experimental designs are
considered classic quantitative methods as they involve manipulating the independent
variable to measure any effects on the dependent variables (Baltimore County Public
Schools, 2017; Creswell, 2014). The remaining types of correlational and descriptive
48
utilize quantitative analysis methods but do not involve any variable manipulation, and so
are considered observational in nature. The study presented here contains both
correlational and descriptive design methods. It is correlational as relationships between
different variables will be examined for and interpreted using statistical analysis methods
(Privitera, 2011). With the collected data, this study will also provide a descriptive
account of how the included participants feel towards and utilize the variables of interest.
In this study, each of the survey items is closed-ended, or focused, in design: they
require each participant to select an answer out of those that are provided and there are no
areas for open-ended, write in responses. The items follow a Likert ordinal scale design.
A Likert scale is used to measure levels of agreement and disagreement in a linear
intensity format (Trochim, 2006a). By ranking the levels of agreement or disagreement
participants may have about a particular concept or statement, it may be possible to
effectively measure the attitudes and beliefs held by a participant pool. All item answers
are assigned a corresponding numerical scale. For example, the Likert scale item answers
have ranking answers that run from 1 being ‘strongly agree’ to 5 representing ‘strongly
disagree’. None of these numerical values mean anything outside of this context: they are
not rankings indicating that one answer is better than another; they are simply numerical
ordinal designations that will allow the data to be evaluated using quantitative analysis
methods.
This quantitative research design that involves closed-ended survey items is
widely used to gather health related data. This method is valuable when investigating for
casual relationships or for trends that may assist in explaining or predicting a health risk
49
or phenomena (Howlett, Rogo, & Shelton, 2014). The collected data is numerical in
value and so can be measured and analyzed. This type of focused data is also objective as
there is no misinterpretation that may occur when examining non-focused, open-ended,
write-in answers. This design is of particular value to this study as it requires minimal
engagement between the researcher and the included participants, thereby allowing data
collection to occur during a quicker timeframe. Also, it allows for a clearer and easier to
understand format that requires only basic comprehension skills in order to complete the
survey. Using a survey with an open-ended question format may have required a higher
set of literacy skills from participants and would have led to longer data collection times
as it may have needed an increased amount of personal interaction between researcher
and participants.
The surveys are paper-based and will be administered by me in person to all
included participants. This method is appropriate to my constraints of time and available
resources, as well as ensures a lower risk of poor response rate, meaning my time will be
more productive than if I had selected a different administration method, such as an
electronic survey.
The survey method of data collection is an integral part of behavioral, social, and
epidemiological research (Saris & Gallhofer, 2007). Questionnaires and surveys can be
specifically tailored to investigate for trends, attitudes and beliefs, or habits in a
population of interest. This is especially valuable when researchers are evaluating for
trends that may have little currently existing documentation. It is for these benefits that I
chose to utilize a survey based research design for my dissertation.
50
Methodology
Population
The Kentucky Appalachian population consists of roughly 1.2 million people who
live across 18,229 square miles (Pollard & Jacobsen, 2011). At 25.4%, the Kentucky
Appalachian area suffers from the highest average poverty rates found in the entire
Appalachian region (FAHE, 2015). Localized pockets across Kentucky Appalachia can
suffer from poverty rates that exceed 40% (Appalachian Regional Commission, 2014).
The poverty rates found in Kentucky Appalachia are greatly increased from the national
average of 15.6%, indicating that a low socioeconomic status is strongly present in this
area (FAHE, 2015). This high poverty rate is partially due to the Kentucky Appalachian
per capita income. In 2014, this per capita income was $30,308, which was significantly
lower than the national per capita income of $46,049 (FAHE, 2015).
Kentucky as a whole suffers from low literacy rates. In 1999, the National Center
for Higher Education Management Systems (NCHEMS) found that 40% of Kentucky’s
working age population had reading skills that fell into the two lowest literacy levels,
those of unable to read and reading at a very limited level (Legislative Research
Commission, 2000). The focus of the report conducted by the NCHEMS was to create
and implement a 20-year strategy to improve the educational and literacy levels found in
Kentucky and so it is unclear if these statistics have changed as the strategy is still in
implementation. As having low literacy is associated with having a low level of health
literacy, it may be assumed that low literacy is contributing to the lower level of health
51
literacy found in the Kentucky Appalachian population as documented by Ludke et al.
(2006).
Low socioeconomic status along with low literacy skills and low health literacy
levels are hallmark indications of both vulnerable and underserved populations. While
vulnerable and underserved are oftentimes used synonymously with each other when
describing populations, the terms actually refer to separate points (Chang et al., 2004).
Vulnerable populations are those that differ from others based on social and demographic
characteristics, such as age, race, or socioeconomic status and who may not properly
utilize available health services (Chang et al., 2004). An underserved population is one
that actually has less than the recommended access to health services due to economic
barriers, or cultural and linguistic differences. Based on the documented risks found in
the Kentucky Appalachian population, it may be possible that portions of this population
may be both vulnerable and underserved in nature, indicating a greater need of
investigation in order to lower the risk of health disparities occurring in this population.
Samples from the Kentucky Appalachian population will be found by polling at
local health departments (LHDs) and churches that are located in the Kentucky
Appalachian region. These local venues are staffed by and serve people who live in the
immediately surrounding communities. They offer population control in this study as it is
unlikely anyone outside of the community of interest would be attending or accessing
these venues.
The included polling venues are in two Kentucky counties, one of which has the
official designation as ‘at risk’ while the other is ranked as ‘distressed’. There is almost a
52
$10,000 a year difference in per capita income between these two counties, indicating
that while both counties represent the region of interest, there is also a clear difference
between the two in economic standing dependent on their county ranking as set by the
Appalachian Regional Commission.
Sampling Procedures
This study will follow a convenience sampling strategy. Convenience sampling is
a type of non-probability method which consists of sampling methods that are based on
the judgement of the researcher, as opposed to probability techniques which are based on
random selection of the included units (Lund Research, 2014). Convenience sampling
involves the inclusion of units that are representative of the population of interest and that
are the easiest to access. This method of sampling oftentimes uses less resources and time
than probability techniques and can allow a researcher to study populations that would
otherwise be difficult to reach (Lund Research, 2014). For this convenience sample,
participants will be selected based on their inclusion status of the local polling venues.
Members of the population that are not accessing the venues at the time of polling will be
excluded from this convenience sample.
Recruitment Procedures
Participants will be recruited by their association with Kentucky Appalachian
LHDs and churches that are located in two different counties. These venues were selected
as data collection sites because they offer a participant control that was lacking in other
data collection sites that were possibly available. These venues serve their immediate
community, which range in size to a few hundred to a few thousand inhabitants. It is
53
unlikely that anyone from outside the immediate communities would have any interest in
the workings of the included local venues, and so the risk of having participants that are
not representative of the area of interest included in this study is low to nonexistent. I
have obtained preliminary permission to poll where I will simultaneously distribute and
collect paper survey-based data. Official permission and appointment dates will be
obtained after IRB approval to do so. The participant pool will include adults conducting
visits at the included LHDs and churches, as well as any adult family/spouses/community
members that may be in attendance with them. The survey itself will be anonymous with
no identifying demographics recorded. There will be no way to link any one set of survey
answers to any one person after survey completion. The projected data collection
timeframe is three to four weeks.
Sample Size
A proper sample size is essential to ensure enough responses to provide accurate
results but to also minimize the risk of too many samples that may use up unnecessary
and valuable resources (Bartlett, Kotrlik, & Higgins, 2001; Smith, 2013). There are many
ways to calculate the projected needed sample size to ensure accurate results, however
these require the reporting of the total population of interest. This brings an ethical
concern to my study in that by reporting the population of my included counties, it would
be then be possible to identify which two counties and locations that were used in this
study. This could lead to possible identification of the included participant pool which
would be a direct violation of my participant’s privacy.
54
When determining a sample size, a researcher needs to take into account many
aspects of their planned study design. These include the research questions and the design
of the study, as well as aspects such as time, available resources, how participants will be
recruited, and projected response rate (Onwuebuzie & Collins, 2007; Scott, n.d.). A
researcher also needs to predetermine statistical guidelines such as level of significance,
the statistical power, and effect size that are being used as these can influence the sample
size as well (Onwuebuzie & Collins, 2007). The study design that I am implementing
involves utilizing local venuess and distributing and collecting surveys in person, so I
anticipate a higher response rate than if I were using an electronic survey or distributing
the survey through the mail system. Time and available resources are definite factors into
my study as both are limited. My statistical guidelines have been set as follows:
Level of significance (α): The probability of rejecting the null hypothesis when it
is true, or a false positive also known as a Type I error. This value is set at 0.05, which is
the value that is most widely used and accepted in research studies (Laerd Statistics,
2013c; Onwuegbuzie & Collins, 2007; Scott, n.d.).
Statistical power: Determined by the value of β which is the probability of failing
to reject a false null hypothesis, or a false negative also known as a Type II error. The
value of β has been set at 20%. With power being calculated by 1 – β, this sets the power
of my study at 80%. This is the probability that I will successfully reject the null
hypothesis. Again, these are very common values that are widely used and accepted in
research studies (Cohen, 1992; Onwuebuzie & Collins, 2007; Scott, n.d.).
55
Effect size: Effect size refers to the magnitude of the difference that may be found
between two groups. This value has been set at 0.5, which is considered to be a moderate
effect size and commonly used (Scott, n.d.; Sullivan & Feinn, 2012).
My study design contains descriptive aspects but is also correlational in nature as
it serves to both gather data that describes how the target population feels about certain
aspects of their dental health as well as examining for any relationships that may exist
between variables of interest. With this study design and the standard statistical
guidelines that have been set, Onwuebuzie & Collins (2007) maintain that 82 samples is
an appropriate sample size for meaningful results. With these findings, my target sample
size is 100, which will encompass the recommendations of Onwuebuzie & Collins (2007)
while staying within my resource limit and protecting the privacy of my participants.
Participation Procedures
Before gathering data, a researcher must first educate their participants on the
fundamental reasoning behind the study as well as obtaining permission from each
participant to include them in the study pool. This process is referred to as informed
consent. Typically, the informed consent consists of a form detailing such points as
explanation of the research purposes, expected duration of the study, along with a
description of the procedures that are included in the study (U.S. Department of Health &
Human Services, 2016). These forms are then signed by the participant as an
acknowledgement that they consent to be included in the study. However, there are cases
where a typically signed informed consent form is not needed in order to conduct a study.
56
Two commonly accepted reasons for not collecting a signed consent form from each
participant include (University of Tennessee, 2017):
1. The informed consent would be the sole record linking the participant to
research data and so could lead to a breach of confidentiality.
2. The research study presents minimal to no risk of harm to the participants.
The study presented here fulfills both requirements as to not use a traditional informed
consent form as this would be the sole record linking the participants to the study, and
being survey based this study presents minimal risks to the included participants. Instead
of the consent form, each participant will be provided a cover letter before being allowed
to take the survey. This cover letter will briefly describe the reason for the study, their
role as a participant, reiterate that their answers are completely confidential, and my
contact information in case there are any questions or concerns that arise after survey
completion. This cover letter will be theirs to keep. Additionally, the first item on the
survey will be ‘I know that I am volunteering to take part in a research study’.
Each participant must check an answer of ‘yes’ to this item in order to continue with the
survey.
Data Collection
Over a data collection period of three to four weeks, at each local venue that is
included, I will personally hand out each survey and be available for any questions
regarding the instructions that each participant may have. A small table will be set up in a
discrete location that will hold extra clipboards with surveys and cover letters attached to
each and pencils. Upon completion, each survey will be immediately placed into a folder
57
for confidential safe-keeping. Participants will exit the study with the completion of the
survey. There will be no follow up with participants after the completion of data
collection. My contact information will be included in the survey cover sheet which will
be also distributed as take home material for all participants. Any questions or
correspondence from participants after taking the survey will be answered and resolved.
Instrumentation
The sole data collection instrument being used is a paper-based survey that has
been developed and specifically tailored for this study. To begin the survey development
process I first determined the research questions and descriptive data that I wished to
include in this study. After these were finalized, I tailored the survey to specifically
answer these research points while making sure I used as plain language as possible so
that the items would be clear and easy to fill out for my population of interest. To keep
the survey as short as possible, many of the items will be used multiple times to answer
all the different research points. For instance, the survey item ‘I floss my teeth every day’
with possible answers of Never, Rarely, Sometimes, Most of the Time, and Always, will
be used in conjunction with other survey items to answer Research Question 1 as well as
Descriptive Item 1:
Research Question 1: Is there an association between the participant’s expressed
level of oral health literacy and the frequency at which they practice oral hygiene
techniques?
Descriptive Item 1: Do Kentucky Appalachian adults’ practice oral hygiene
techniques at frequencies as recommended by the American Dental Association?
58
So in this example, the same information will be used to assist in answering how often
Appalachian adults are utilizing oral hygiene methods, as well as investigating on if the
usage rate changes with their expressed level of oral health literacy.
Reliability
To provide evidence for the reliability of the survey, the method of internal
consistency reliability will be utilized. Internal consistency reliability is the method of
using multiple versions of the same survey question in order to determine if there is a
consistency between the two answers (Trochim, 2006b). For instance, on the survey used
in this study, the following two items are asking about the same information just in
differing formats:
1. I brush my teeth two times every day.
2. Sometimes I don’t brush my teeth every day.
Internal consistency reliability would be demonstrated if these two items pulled similar
answers: for instance, someone who answered as ‘always’ for item one should then
answer ‘never’ to the second item to demonstrate appropriate consistency. Internal
consistency can further assist in demonstrating that the data collection instrument is
working appropriately.
Validity
Construct validity describes how well a test actually measures for a particular
concept (Trochim, 2006b). In this presented study, the construct being measured would
be the level of oral health literacy that each participant expresses by way of answering
survey items. The validity here would be in how well the survey items are measuring the
59
construct of oral health literacy. To measure the construct validity in this study, the
method of face validity can be utilized. Face validity is the evaluation of the data
collection instrument and deciding if it will work to collect that measures the concept in
question (Trochim, 2006c). To do this, the survey will be sent to a selected sample of
experts to determine if they agree that the instrument appears to be measuring the
construct of oral health literacy correctly. By including multiple experts, such as those
who have established experience in research data collection, this ensures that the
instrument must gain approval from multiple venues before it is used.
Operationalization
Independent variable: Oral health literacy. This is defined by the level of oral
health literacy that is expressed by participants as determined by how they answer survey
items. The items that are being used to evaluate for oral health literacy will provide
continuous data for examination. The answers will be ranked on a Likert scale as follows:
Strongly Disagree = 1
Disagree = 2
I Don’t Know = 3
Agree = 4
Strongly Agree = 5
Dependent variable: Brushing and flossing frequencies. This is how often
participants report that they brush and floss their teeth. This variable will be determined
by how participants answer survey items. The items used to determine this variable
60
utilize the same Likert scale as described above, and also a second Likert scale as
follows:
Never = 1
Rarely = 2
Sometimes = 3
Most of the time = 4
Always = 5
All remaining variables will be measured using these Likert scale formats.
Data Analysis Plan
The software that will be used in the analysis of the collected data for this study is
the widely utilized Statistical Package for the Social Sciences, or commonly referred to
as SPSS. SPSS is a program produced by IBM that will allow for the organization and
analytic testing needed to identify any trends in the collected data (IBM Analytics, n.d.).
The data collected on the paper surveys will be transferred by me into the SPSS program
in order to create the electronic database. During this process, any survey that is found to
be incomplete will be eliminated, thereby ensuring a clean and proper final database.
Restatement of Research Questions and Descriptive Items
Research Questions:
1. Is there an association between the participant’s expressed level of oral health
literacy and the frequency at which they practice oral hygiene techniques?
Ha1: There is an association between the participant’s expressed level of oral
health literacy and the frequency that they practice oral hygiene techniques.
61
H01There is no association between the participant’s expressed level of oral
health literacy and the frequency that they practice oral hygiene techniques.
2. Is there an association between the participant’s expressed level of oral health
literacy and the frequency at which their children practice oral hygiene techniques?
Ha2There is a relationship between the participant’s expressed level of oral health
literacy and the frequency that their children practice oral hygiene techniques.
H02:There is no relationship between the participant’s expressed level of oral
health literacy and the frequency that their children practice oral hygiene
techniques.
3. Are Kentucky Appalachian adults and children practicing oral hygiene techniques at
the same frequencies?
Ha3:There is a difference in how often oral hygiene techniques are used between
adults and children.
H03:There is no difference in how often oral hygiene techniques are used between
adults and children.
4. Were there differences in the survey responses gathered from the transitional
Kentucky Appalachian county and the distressed Kentucky Appalachian county?
Ha4:There are differences in the survey responses collected between the
transitional and distressed counties.
H04: There are no differences in the survey responses collected between the
transitional and distressed counties.
Descriptive Items:
62
1. Do Kentucky Appalachian adults practice oral hygiene techniques at frequencies
as recommended by the American Dental Association?
2. Do Kentucky Appalachian children practice oral hygiene techniques at
frequencies as recommended by the American Dental Association?
3. Do Appalachian adults perceive dental decay as a health risk?
4. Do Appalachian adults feel that dental decay is a preventable health risk?
5. Do Appalachian adults feel that they can personally decrease the risk of dental
decay?
6. Do Appalachian adults perceive oral hygiene techniques as important for good
health?
7. To what level do Appalachian adults perceive that enforcing childhood oral health
techniques will decrease the risk of dental decay as their children age?
8. Do Appalachian adults perceive poor dental health as a normal event?
Analysis Plan
Traditionally, Likert scale data has been considered to be ordinal in nature in that
the answers can be ranked but the actual distance between the answers cannot be
measured as they can be highly subjective between participants (Sullivan & Artino,
2013). For example, the distance between answers of ‘completely disagree’ and
‘disagree’ cannot be measured as the exact distance because the meaning of each answer
can be different between the included participants. For the purposes of this study, the
Likert scale data being collected will be treated as ordinal and so will have non-
parametric tests applied for analysis. Non-parametric tests are those that do not make
63
assumptions about the distribution, unlike parametric tests that assume a normal
distribution in the population of interest (Frost, 2016; Sullivan & Artino, 2013). As such,
parametric tests such as means and standard deviations will not be computed, but instead
I will utilize nonparametric tests such as median, frequencies, and the Mann-Whitney U
test to answer the research questions and descriptive items.
To begin the data analysis, I will transfer the collected data into SPSS thereby
making an electronic data table. The first calculation that will be performed on the newly
created electronic data set is to compute for Cronbach’s alpha. In 1951, Lee Cronbach
created the Cronbach alpha in order to measure a scale test’s internal consistency
(Tavakol & Dennick, 2011). Internal consistency is how well all the items included on a
test measure a concept or attitude (Tavakol & Dennick, 2011). The better the items on a
test demonstrate connectivity, or inter-relatedness, the higher level of validity the test
may be said to have. The Cronbach’s alpha is expressed by a number that falls between 0
and 1, with values above 0.65 being acceptable (Tavakol & Dennick, 2011; University of
Virginia Library, 2017). If the included scale data items are independent of each other,
that is, they have no inter-relatedness and share no covariance, then the Cronbach’s alpha
equals 0 and the test is considered to have no internal consistency (University of Virginia
Library, 2017).
Numerically, Cronbach’s alpha is defined as:
α = k x c̅
v̅ + (k - 1)c ̅
64
Where alpha includes the average covariance between pairs of items as well as the
variance of the total score (University of Virginia, 2017). This calculation will be done
using SPSS to reduce the risk of error.
After the data has been tested for inter-relatedness, each survey item will have its
median answer calculated by including all corresponding item answers. For instance, all
the answers for item 2 will be included in the calculation and this will give a median
answer for item 2. The median is the nonparametric equivalent to mean and will give the
midpoint answer for each item. As previously discussed, in this study each research
question has two or more items associated with it in order to provide more internal
consistency. To answer the descriptive items, the composite median score of the included
items will be computed. For instance, to answer research question 5:
‘Do Appalachian adults feel that they can personally decrease their risk of dental
decay?’, the medians for three survey items will be included. The composite median of
these three median scores will be the answer to research question 5. By computing the
median value for each survey item, this will provide information as to the average
participant response to each item (Kostoulas, 2014).
Next, the interquartile range (IRQ) will be calculated for each survey item. The
IQR measures how the middle 50% of survey responses are dispersed and will show if
the responses are clustered around one answer, or if they are scattered across the possible
answers (Kostoulas, 2014; University of Leicester, 2017). To calculate the IQR the
responses for each survey item will be arranged in a ranked-order format, similar to how
the data is arranged by magnitude in order to compute the median value. The ordered
65
responses will then be divided into four equal parts (University of Leicester, 2017). The
values that separate each part are referred to as quartiles and these are the values that will
be used to calculate the IQR as seen in the following equation:
IQR = quartile 3 – quartile 1
A smaller IQR value will indicate that the responses are more clustered around a
particular answer, thereby showing a more unified consensus among the participant pool
(Kostoulas, 2014). However, if the IQR is a larger value, this could suggest that the
participants have strong opinions both for and against the survey item (Kostoulas, 2014).
The IQR will assist in determining if the median value is an accurate representation of
average opinions of the participant pool. A small IQR that is suggesting consensus among
the answers supports the median value as an accurate report of how the participants as a
whole felt about a particular survey item (Kostoulas, 2014). Conversely, a large IQR
value that suggests a wider array of strong feelings both for and against the survey item
indicates that the median value is likely not an accurate indicator of the average reported
response.
To answer the listed hypotheses, a more in-depth analysis method must be used as
each hypothesis requires that a correlation be investigated for between a dependent and
independent variable. If a correlation between the two variables can be established, then
the null hypothesis is rejected, thereby providing an answer to each hypothetical query.
To do this, the Spearman rank-order correlation, also referred to as Spearman’s
correlation, will be used. The Spearman correlation is a nonparametric test that measures
the magnitude and direction of an association that is suspected to exist between two
66
ordinal variables (Laerd Statistics, 2013a). The Spearman correlation test produces a
correlation coefficient that can determine the strength of the association between the two
variables of interest. A correlation coefficient of .50 and above is read as a strong
association between variables, with .29 and below being considered a weak association.
The previously discussed tests of Cronbach’s alpha, median values, interquartile
range, and Spearman’s correlation will all serve to provide information on the attitudes
and habits the included Appalachian participants hold towards dental health and care, and
thereby answering the main research questions and descriptive items. A further test will
be done that evaluates for any differences in survey responses based on the location
where each survey was completed. In this study, surveys are being offered in a
transitional Appalachian county (County A) and a distressed Appalachian county (County
B). To complete these calculations, the Mann-Whitney U test will be utilized. The Mann-
Whitney U test is a nonparametric test that is comparable to the parametric independent
samples t-test and that can compare and identify differences between two independent
groups (Laerd Statistics, 2013b). For the Mann-Whitney U test, the data from each survey
item will serve as the ordinal scale dependent variables, while County A and B
designations will serve as the independent grouping variables. The Mann-Whitney U test
calculations will be completed by SPSS. These calculations will provide a Ranks table
that will display the mean rank and sum of ranks between the two groups of participants
(Laerd Statistics, 2013b). This will show if there are differences per survey item based on
county designation. SPSS will also generate a Test Statistics table as part of the Mann-
Whitney U test function. This table will provide the U statistic and the p-value, both of
67
which are used to determine if the mean ranks are significantly different from each other
(Laerd Statistics, 2013b). If there are significant differences found, this implies that
people are answering an item differently based on county designation.
Threats to Validity
External Validity - Threats
External validity refers to how well the findings of a study can be applied and
used as generalizations towards other people, places, and times that were not included in
the study (Steckler & McLeroy, 2008; Trochim, 2006d). Threats to external validity are
any that reduce a study’s generalizability of its results (Laerd Dissertation, 2012a).
External validity threats are found as two specific categories: ecological validity and
population validity (Andale, 2016; Michael, n.d.). Ecological validity is how well the
results can be generalized across settings or places that were not included in the study.
Most types of ecological validity threats involve the use of pre-testing or education that
may interfere with how participants respond to the study’s experimental treatment,
thereby reducing the generalizability of the results to a population that does not have the
pre-test or education (Andale, 2016; Michael, n.d.; University of Minnesota, 2017).
These types of threats will not affect the validity of my study as no pre-testing or
education is being provided before data collection. However, ecological validity also
includes the threat of reactive effects of experimental arrangements, also known as the
Hawthorne effect. The Hawthorne effect describes how the knowledge of participating in
a research study may impact the answers or behavior of the study participants
(McCambridge, Witton, & Elbourne, 2014). Knowing that they are being researched may
68
influence a participant to answer how they think they should answer as opposed to
providing an answer that more accurately portrays their opinion or feelings on a subject.
My study will be at risk from the Hawthorne effect as it is impossible for my included
participants to not know that they are taking part in a research study and so this will be
reported as a threat against the generalizability of my findings.
Population validity refers to the extent that the included study participants are
accurate representations of the population of interest (Ferguson, 2004; Michael, n.d.). An
appropriate participant pool is essential for generalizable results as population validity is
seen as a key threat to the overall external validity of a study (Ferguson, 2004). The
random selection of participants is largely viewed as the best method of obtaining a
sample of participants that represent the population of interest and produces the highest
result generalizability, particularly in quantitative styled studies (Ferguson, 2004;
Onwuegbuzie & Collins, 2007). However, Leech & Onwuegbuzie (2002) maintain that in
actuality, the majority of quantitative studies utilize non-random sampling techniques and
that these non-traditional methods allow researchers more options and opportunities when
selecting study participants. For my study, participants are included based on their
attendance of local venues that are located in counties that are designated as Appalachian.
The benefits of using these venues are as follows:
1. These venues are located in the area of interest.
2. These venues do not bring bias to the study such as would happen if for example,
county fairs were used as data collection points.
69
3. These venues are frequented by people who live and work in the immediate area
that the venues serve, thereby allowing me to reach the population of interest,
even in rural and more remote areas.
4. Using these venues will allow for direct access to Appalachian adults while
providing assurance that people from outside the area of interest will not be
involved in the study as it is unlikely that anyone outside the immediate service
district will use these venues.
Other than the Hawthorne effect, there are very little to no other threats to the external
validity of my study. This means that my results may be able to be used as
generalizations for similar people or places that were not included in this study.
Internal Validity - Threats
While external validity refers to the generalizability of the results beyond the
scope of the study, internal validity is concerned with how well the concepts of interest
are actually measured, particularly in studies where associations or relationships are
being established (Kimberlin & Winterstein, 2008; National Business Research Institute,
2017; Trochim, 2006e). Another way of describing internal validity is that it is the extent
to which a study’s results are related to the independent variable of interest, as opposed to
some other variable. Internal validity is closely related to the reliability of a study
(Kimberlin & Winterstein, 2008). However, while validity requires that a data collection
instrument such as a survey be reliable, the instrument can be reliable without being valid
(Kimberlain & Winterstein, 2008).
70
Threats to internal validity are mainly a concern with studies that are investigating
for cause-effect relationships and are usually not relevant in observational or descriptive
studies (Trochim, 2006e). The study presented here follows a primarily descriptive
format in that its prime focus is to evaluate for the opinions and attitudes health towards
oral health and the participants are being polled only one time with no experimental
treatment. As such, most of the types of internal validity threats such as maturation,
mortality, and contamination effect are not applicable to my study. However, two types
of threats could exist in my study. The first of these is referred experimenter effects or
bias (Laerd Dissertation, 2012b). This type of threat to internal validity occurs when the
researcher somehow influences the choices made by the participants during the study and
so makes the resulting data biased and unreliable (Laerd Dissertation, 2012b). Even if
these researcher effects are unintentional, they can still lead to incorrect results. While
experimenter or researcher effects are more likely to be found in qualitative research that
may involve longer face-to-face interviews between researcher and participant, I still
need to ensure I reduce the risk of this threat in my quantitative study. To do this, I need
to be clear about the survey directions and the point that there are no wrong answers to be
found in the survey, that it is indeed strictly measuring only opinions. I also need to be
careful when addressing each participant so that I don’t unintentionally lead to them to
answers they think I may be looking for to answer my research questions. To accomplish
this, I believe it will be most effective to not speak of the contents of the survey with any
of the participants. For example, if I am asked for direction clarification, I will use an
example question that is unrelated to oral health to explain the procedure.
71
Another threat to the validity of my study is subject effects, which is also referred
to as participant reactivity (Laerd Dissertation, 2012b). Participant reactivity is similar to
the previously discussed Hawthorne effect, but deals more with how participants will
actually change their behavior when taking part in a research study. This change in
behavior could possibly influence how participants respond to my survey. However, I am
reducing the threat of subject effects by surveying my participants in an environment that
is familiar and less staged for them. If I were to conduct my survey in a lab or a location
that is new to my participants, I would increase the risk of behavior change in my
participants because it may increase the feelings of being scrutinized, examined, and
unease towards being in an unfamiliar location (Laerd Dissertation, 2012b).
Construct – Threats
Out of the threats that can occur against the construct validity of a study, my study
may be at risk of inadequate preoperational explication of constructs. More simply, this
term refers to how an idea or concept that is being evaluated as a construct has not been
properly defined to adequately explain what the researcher means (Strauss & Smith,
2009). To avoid this threat to the construct validity of this study, I have been clear in my
definitions of the concepts that are included in my study. These include the independent
and dependent variables, as well as conceptual ideas and theories such as oral health
literacy.
Another threat to the construct validity of this study falls under the category of
mono-method bias. Mono-method bias is when bias in introduced into a study’s results
due to using only one measurement method (Laerd Dissertation, 2012c; Strauss & Smith,
72
2009). In this study, I attempt to discover the current dental health status of my
participants by using survey items designed to develop a self-perceived dental health
status of the included participants. This may introduce bias into my study as this is the
only form of dental health status establishment that I am including. If I were to utilize
multiple methods of data collections, such as survey items along with dental health
records, I may end up with a very different overall dental health status of the participant
pool than by just using survey items alone. However, it is beyond the scope of this study
to include any other methods, such as dental health record collection and analysis, other
than the survey instrument. Mono-method bias will be reported in the results as a possible
threat to the construct validity of this study.
Statistical Conclusion Validity – Threats
Statistical conclusion validity is how well data can be regarded as accurately
identifying an association, or lack of, between independent and dependent variables (as
reported by Garcia-Perez, 2012). In a quantitative study, the statistical conclusion validity
is threatened when inadequate statistical analysis methods are conducted on the results
(Garcia-Perez, 2012). Not using enough variety of or improper statistical analysis
methods can yield results that are not accurate (Garcia-Perez, 2012). To reduce the threat
of statistical conclusion validity, researchers should thoroughly examine a wide variety of
statistical analysis methods and select to use as many that are deemed appropriate in
order to fully analyze the resulting data (Garcia-Perez, 2009; Milligan & McFillen,
1984). After much researching of survey-based data analysis methods, I have selected to
use multiple analysis methods to include Cronbach’s alpha, median values, answer
73
frequencies, Spearman correlations, and the Mann-Whitney U test. By utilizing as many
appropriate statistical analysis methods as possible, my data will be thoroughly analyzed
which will reduce the threat to the statistical conclusion validity and also reduce the risk
of Type-I and Type-II errors in my results.
Ethical Procedures
Institutional Permissions
IRB approval through Walden University was obtained before data collection was
conducted in this study. The IRB number for this study was 12-29-17-0078647.
Ethical Concerns
The main ethical concern that has been present in this study is keeping the identity
of the participants anonymous. This study will mimic other studies that have conducted
research in the Kentucky Appalachian region in that the included county names or exact
polling locations will not be named in this dissertation (Dawkins et al, 2013; Ludke et al.,
2006; Savage et al., 2014).
One of the statistical analysis methods that I will utilize on the resulting data
involves comparing the survey answers from County A to the results of County B. To
achieve this, each survey will be numbered and I will keep a record of which survey
numbers were handed out in each county. This will be the only use of the survey numbers
and they will not be linked to the actual participants in any way. This record will only be
seen and utilized by me and I will keep it along with other sensitive study information.
74
Data Type and Storage
Due to my plans of comparing data between the two counties based on their
economic rankings, it will be possible for me to identify which survey came from each
county. However, it will be impossible to tie any one survey to a single person and
therefore my data is anonymous in nature. This assurance will assist in alleviating any
concerns my participants may have about the protection of their privacy when responding
to the survey.
Data collection will be completed by using paper surveys. The data will be
transferred into SPSS in order to create an electronic data base that can then be used for
analysis. The paper surveys will be kept in a secured and locked drawer of my desk that
is located in my home. Only I have access to the key and only I will have access to both
the paper surveys as well as the electronic data base. These paper surveys will be kept
secure for a period of five years after which they will be destroyed by being mechanically
shredded and offered for recycling. The electronic data base will be strictly accessed only
by me and only from my home desktop computer, which is password protected and
where I am the only user. No other people will have access to the data contained in the
electronic database. All documents pertaining to my dissertation, including the electronic
database created from the survey responses, will be kept saved on my personal hard drive
with a back-up copy saved on my personal Google Drive. The electronic version of my
dissertation will be kept indefinitely on my personal home computer and my Google
Drive.
75
Compensation
There is no compensation being offered to people who chose to participate in this
study.
Summary
The presented study utilizes a quantitative study design as numerical data will be
collected. The data will be gathered by using paper-based surveys that contain items to be
answered by closed-ended Likert scale options. The electronic Likert scale data set will
be analyzed using several statistical techniques in order to provide answers to the
research questions, hypotheses, and to identify any outstanding trends that may emerge
from the included participant pool. Participants will be invited to take part in a drawing
which will provide compensation at a rate that cannot be considered coercion to take part
in this study. No one participant will be able to be identified with any one particular
survey response: the anonymity of the study is protected. All data will be kept securely in
my home where only I will have access to any records, including the paper surveys and
resulting electronic data base. The electronic version of all data pertaining to this
dissertation will be kept indefinitely on my own personal, secured computer hard drive
and backed up on my personal; pass-word protected Google Drive
76
Chapter 4: Data Analysis
Introduction
The purpose of this study was to evaluate for a relationship between the overall
levels of expressed oral health literacy and how often people in Kentucky Appalachian
areas utilize oral hygiene techniques. Additionally, this study also serves to give a
descriptive overview into how people in Kentucky Appalachian areas may feel towards
dental health and how it pertains to them. At the time of this writing, little to no
established data could be found that serve to describe or document the attitudes or beliefs
that Appalachian Kentuckians hold towards these concepts. Lastly, an investigation is
conducted to determine if there are differences in responses based on their national
county ranking. As previously discussed in the literature review section of this study,
counties across the United States hold a national ranking position that is used to indicate
each county’s economic status. Counties holding the rank of transitional are considered to
be transitioning from weak economies to strong, or vice versa, and make up the worst 25
to the best 25 percent of the nation’s counties. Counties that are ranked as distressed
make up the worst 10 percent of the nation’s counties and so indicate that these are the
most economically distressed areas. The study presented here included participants from
one distressed and one transitional Kentucky Appalachian county.
As a review, the following research questions, hypothesis, and descriptive items
have been developed. Details of how the included variables and concepts were
operationalized, as well as the statistical analysis used for each inquiry will be provided
later in this chapter.
77
Research questions:
1. Is there an association between the participant’s expressed level of oral health
literacy and the frequency at which they practice oral hygiene techniques?
Ha1: There is an association between the participant’s expressed level of oral
health literacy and the frequency that they practice oral hygiene techniques.
H01: There is no association between the participant’s expressed level of oral
health literacy and the frequency that they practice oral hygiene techniques.
2. Is there an association between the participant’s expressed level of oral health
literacy and the frequency at which their children practice oral hygiene techniques?
Ha2: There is a relationship between the participant’s expressed level of oral
health literacy and the frequency that their children practice oral hygiene
techniques.
H02: There is no relationship between the participant’s expressed level of oral
health literacy and the frequency that their children practice oral hygiene
techniques.
3. Are Kentucky Appalachian adults and children practicing oral hygiene techniques at
the same frequencies?
Ha3: There is a difference in how often oral hygiene techniques are used between
adults and children.
H03: There is no difference in how often oral hygiene techniques are used
between adults and children.
78
4. Were there differences in the survey responses gathered from the transitional
Kentucky Appalachian county and the distressed Kentucky Appalachian county?
Ha4: There are differences in the survey responses collected between the
transitional and distressed counties.
H04: There are no differences in the survey responses collected between the
transitional and distressed counties.
Descriptive Items:
9. Do Kentucky Appalachian adults practice oral hygiene techniques at frequencies
as recommended by the American Dental Association?
10. Do Kentucky Appalachian children practice oral hygiene techniques at
frequencies as recommended by the American Dental Association?
11. Do Appalachian adults perceive dental decay as a health risk?
12. Do Appalachian adults feel that dental decay is a preventable health risk?
13. Do Appalachian adults feel that they can personally decrease the risk of dental
decay?
14. Do Appalachian adults perceive oral hygiene techniques as important for good
health?
15. To what level do Appalachian adults perceive that enforcing childhood oral health
techniques will decrease the risk of dental decay as their children age?
16. Do Appalachian adults perceive poor dental health as a normal event?
79
The remainder of Chapter Four will discuss the process of data collection, results of each
research question and descriptive item, and will finish with a summary of the main trends
that were identified in the collected data.
Data Collection
As previously discussed throughout the proposal of this study, the protection of
the identity of the included participants was an utmost priority. This can be difficult to
achieve when using very rural polling venues such as were used in this study. After
consideration, it was decided that actual county names or polling venue names would not
be included in this study. This follows the privacy methods utilized by Ludke et al.
(2006), Dawkins et al. (2013), and Savage et al. (2014). These researchers all conducted
studies in Kentucky Appalachian areas and omitted any references to exact areas or
locations in which their data was collected. To this end, the transitional county included
in this study is referred to as County A and the included distressed county is designated
as County B. This fully protects the identity of the included participants, polling venues,
and the officials who gave permission to poll at the venues.
The time frame of data collection stretched over a span of roughly four weeks.
Data collection in County A took place over a span of three weeks in which two different
events were attended and were used as data collection venues. Data collection in County
B took place over two days in which one specific community venue was used for polling.
One discrepancy from the data collection plans presented in chapter three was that
local fire departments (FDs) were not able to be obtained as polling venues, largely due
to a lack of interest in local FD staff. Instead, local health departments (LHDs) and
80
churches were used as polling venues. These venues presented the same population
control parameters as local FD’s as both LHD’s and churches are mainly utilized by
people in the more immediate surrounding area. By choosing venues located in areas of
interest, i.e. counties that are designated as transitional and distressed, I was able to
ensure I did not include participants that were not representative of my population of
interest.
Although the actual polling places had to be changed, the fundamental concept of
comparing responses from a transitional county against the responses gathered in a
distressed county did not. As of the Fiscal Year 2018 County Economic Report published
by the Appalachian Regional Commission, there are three Kentucky Appalachian
counties that have the rank of transitional (Appalachian Regional Commission, 2018).
The remaining 51 counties that make up the Kentucky Appalachian region are ranked as
at-risk or distressed: no counties in this area have the highest rankings of competitive or
attainment. County A that was included in this study is one of the three counties with the
ranking of transitional. This ranking indicates that County A has a higher level of average
economic status and education but that there may be pockets of people present that may
be living at economic levels lower than the national poverty level. County B that was
included in this study has the ranking of distressed, indicating that people in this county
are living in one of the country’s most economically poor areas. The demographics of
both counties included in this study were investigated to ensure they fully fit the criteria
of this study. Although exact numbers cannot be reported here due to the possibility of
identifying the included counties and participants, particularly as there are only three
81
counties in the transitional status, general demographics can be remarked upon to
demonstrate that the included counties are representative of the population of interest.
County A was found to have over four times the population of County B, however nearly
40% of the population of County B was found to be enrolled in Medicaid, compared to
County A which was under 20% (Kentucky Health Facts, 2018). The median household
income of County A is higher than County B by nearly $15,000. The estimated median
income between County A and County B was $31,300. Twenty-six percent of the
respondents from County A reported a yearly income below $31,300 while 41% of the
respondents from County B reported the same. County A has nearly five times the
number of practicing dentists than that of County B. County A has an 11% increased high
school graduation rate over County B. These general demographics all demonstrate that
the included counties are representative of the Kentucky Appalachian area of interest as
defined by the included counties national rankings of transitional and distressed.
To obtain participants from the included counties, the convenience sampling
method was utilized. This included the polling of naturally occurring groups of people
visiting selected venues located in the included counties. Population control was supplied
by only selecting venues that weren’t likely to be utilized by people outside of the area of
interest, such as churches and local health departments. While the entire population did
not have the chance to participate in this study, the convenience sampling method
allowed for obtaining participants that are representative of the larger population of
interest (Research Methodology, 2018).
82
Results
In this section, the results of the data analysis are presented. First background
behind the statistical analysis tests that were used is discussed, and next the computed
data is presented with a discussion of the results following. Effort was made to ensure
that multiple Likert type items asked about the same factor were included instead of
relying on a sole item to answer any particular research question. In this way, a Likert
scale construct was created for each factor of interest.
This study involved the use of Likert type survey items where the participants
could choose the answer that best fit their opinion of each statement. At the time of this
study, no previously developed questionnaire designed to investigate the concepts of
interest held by this study could be found for use. As such, a questionnaire was developed
specifically for use in this study (See Appendix for the survey items that were developed
and used for data collection).
There is a long standing, continuing debate between using parametric or non-
parametric methods when analyzing Likert type items or scales (Carifio & Perla, 2008).
This largely stems from the differing opinions on whether or not Likert type items and
Likert scales should be treated as ordinal or interval data (Allen & Seaman, 2007; Carifio
& Perla, 2008; Murray, 2013). For the purposes of this study, the Likert type items and
resulting Likert scales were treated as ordinal data and thus non-parametric analysis
methods were utilized. This choice was largely made due to this study involving a survey
in which people recorded their opinions on various items and so the distributions may not
83
have been normal, which is an assumption of parametric testing methods. Non-parametric
testing methods make no assumptions about the distributions.
Being non-parametric in nature, the Cronbach’s alpha value was included as it is
the most widely and frequently used method of testing the internal consistency of the
Likert type items included in each Likert scale (Gadermann, Guhn, & Zumbo, 2012;
Tavakol & Dennick, 2011). Upon calculation, most of the Cronbach’s alpha values were
found to be subpar, i.e., below what is considered the acceptable value range of 0.60 to
0.90 (Tavakol & Dennick, 2011). After further investigation, it was discovered that
Cronbach’s alpha isn’t always a dependable measure of a scale’s internal consistency and
can lead to misinterpretation of or even wrongly discarding survey results (Sijtsma, 2009;
Tavakol & Dennick, 2011). The Cronbach’s alpha value decreases as the skewness of the
scale items increases (Gadermann et al., 2012). This means that for survey scales that do
not have a clear consensus of answers, meaning that participant responses are scattered
among all the answer choices, the alpha will decrease. Many of the Likert type items and
thus the resulting Likert scales included on the survey used in this study obtained
scattered response patterns, indicating skewness and thus producing lower Cronbach
alpha values. This drawback of Cronbach’s alpha was demonstrated when calculating the
value for Descriptive Item 2: ‘Do Kentucky Appalachian children practice oral hygiene
techniques at frequencies as recommended by the American Dental Association?’. The 28
responses that indicated that the participant had no children were removed from the data
table and a Cronbach’s alpha value of .544 was obtained from the remaining 71
responses. Out of curiosity the 28 removed responses were added back into the data table
84
and the resulting Cronbach’s alpha jumped up to .921.On face value, the second alpha of
.921 would be much more acceptable, however, it is taking into consideration the 28
participants who responded the same with no children. In actuality, including these 28
participants changed the distribution of the answers in a way that is not truly
representative of the data: participants who responded as having no children should not
be included in items that ask about children’s oral hygiene habits. An additional trend
was identified in items that had IQR values of 1.5 or higher, indicating more uneven
distribution in the responses, were the items that had the lowest Cronbach’s alpha values.
This further supports that Cronbach’s alpha may not be an appropriate internal
consistency measurement when evaluating an opinion survey. Although Cronbach’s
alpha was possibly found to be a non-reliable method of measuring for internal
consistency in this study, the values are still reported for each appropriate item. This is
due to the fact that Cronbach’s alpha is still the most frequently used value and the
alternatives for assessing reliability aren’t as well known, nor were they found to be
readily calculable using available software such as SPSS. Instead, lower values of
Cronbach’s alphas were accepted and much focus was given to the ratio of answers for
each survey item to identify any trends.
Research Questions and Descriptive Items Data Analysis
In this section the statistical analysis is shown for each of the research questions
and descriptive items included in this study. It is divided by each research question and
descriptive item with all included survey data for each being listed and the corresponding
analysis immediately following (refer to Appendix for the survey items that were used in
85
this study). For analysis purposes, all survey item answers were assigned a numerical
value. For instance, the survey item statement ‘I floss my teeth every day’ had numerical
scores assigned to each of the answer choices as follows: 1 = Never, 2 = Rarely, 3 =
Sometimes, 4 = Most of the time, 5 = Always. A participant choosing to answer this
statement with ‘sometimes’, had a score of 3 recorded for this survey item. In this way
each participant’s responses were converted to numerical data for analysis and allowed
for the ordinal data to be ranked as the numerical values imply a ‘greater than’
relationship. For the corresponding reverse survey items, the scores were reversed. From
the example used above, the corresponding reverse survey item was ‘Sometimes I am just
too busy to floss my teeth every day’. The answer choices for this item were scored as
follows: 1 = Always, 2 = Most of the time, 3 = Sometimes, 4 = Rarely, 5 = Never. In this
way, each concept that was investigated did not rely on a sole survey item and the scores
for each item correctly aligned to the response choices. Single Likert items used alone are
considered to be less valid and less reliable than Likert scales that are composed of
multiple items, particularly when measuring perceptions people may hold towards a
concept of interest (Warmbrod, 2014).
As demonstrated, multiple survey items were utilized in order to create Likert
scales that were specifically designed for each inquiry contained in this study. A total
composite score for each participant was calculated by finding the median of the
numerical values for all of their responses to the survey items included in each Likert
scale. For each separate Likert scale used in this study, care was taken to ensure that all
included Likert-type items had the same range of answer choices. For instance, when
86
creating the Likert scale for investigating how often participants utilized oral hygiene
methods, only items that had answer choices of ‘never’ ranging to ‘always’ were included
in the Likert scale. No items with answer choices of ‘strongly disagree’ ranging to
‘strongly agree’ were used. This allowed for the total composite scores to be inferred the
same way as the individual Likert-type items: a total composite score of 4 or 5 indicated
that that participant utilized oral hygiene methods more often than participants with total
composite scores of 1 or 2. The use of total composite median scores also allowed for the
reporting of the central tendency of the answer responses where appropriate throughout
the data analysis process.
Presented first are research questions one and two, in which similar statistical
analysis methods were utilized in order to answer each by way of rejecting or failing to
reject each associated null hypothesis. Lastly, Cronbach alpha values and Spearman’s
correlation coefficients were computed for each question. Considering these results in
their entirety allowed for answers to be obtained for the first two research questions as
follows.
Research Question 1: Is there an association between the participant’s expressed
level of oral health literacy and the frequency at which they practice oral hygiene
techniques?
Ha1: There is an association between the participant’s expressed level of oral
health literacy and the frequency that they practice oral hygiene techniques.
H01: There is no association between the participant’s expressed level of oral
health literacy and the frequency that they practice oral hygiene techniques.
87
To determine the variable of expressed oral health literacy, multiple survey items
that investigated for several concepts were included. These included concepts were how
Kentucky Appalachian adults perceive dental decay as a health risk, if they felt dental
decay is a preventable health risk, and if they felt they could personally decrease the risk
of dental decay. While no previously established measurement for oral health literacy
level could be found, these concepts were included in this study as they mirror concepts
that are documented as contributing to a person’s overall health literacy level (Helitzer,
Hollis, Sanders, & Roybal, 2012). The survey items pertaining to how participants
perceived dental decay as a health risk were:
Having poor teeth can lead to other health problems.
Having poor teeth can make me sick in other ways.
The included items that investigated for if participants felt that dental decay is a
preventable health risk were:
If parents have poor teeth their kids will too.
If I brush and floss, I will have good teeth.
If kids floss their teeth while they are young, they will have better teeth as
adults.
If kids brush their teeth while they are young, they will have better teeth as
adults.
And finally, to include if the participants felt they could personally decrease the risk of
dental decay the following survey items were included:
I can only have good teeth if I see a dentist often.
88
Kids can only have good teeth if they see a dentist often.
The median values of the responses to each of the above listed eight survey items that
made up the Likert scale for oral health literacy were calculated, thereby giving a total
composite median score for each participant. These total composite median scores then
represented the level of oral health literacy that was expressed by each participant by
considering their responses to the included concepts of interest. Using the total
composite median scores for each participant produced values that fell between the
ranges of 1 and 5. Keeping in that all the included Likert-type items had the same
response choices and that ordinal data that has been assigned numerical values implies a
‘greater than’ relationship, it is inferred that a participant who had a total composite score
of 5 expressed a higher oral health literacy level than a participant with a total composite
score of 1 (Clason & Dormody, 1994; Warmbrod, 2014)..
After establishing the expressed level of oral health literacy for each participant,
the frequency at which each participant utilizes oral hygiene techniques was calculated.
The following survey items were included to investigate the usage of brushing and
flossing:
I floss my teeth every day.
I brush my teeth twice every day.
Sometimes I am just too busy to brush my teeth twice a day.
Sometimes I am just too busy to floss my teeth twice a day.
Again, a total composite median score was calculated for each participant that included
the responses from each of the four included survey items. It is of note that for the
89
purposes of this study, the definition of proper oral hygiene usage is considered to consist
of brushing twice per day and flossing once per day as recommended for proper results
by the American Dental Association (2016). Any differences in the reported rates
between brushing and flossing are covered later in this chapter. The Cronbach’s alpha
scores for each construct are as follows:
Table 2
Cronbach's Alpha Scores for Research Question 1
Perceived Risk Perceived as Preventable Perceived Self
Control
Adult
Brush/Floss
0.424 0.522 0.574 0.571
The omission of any of the Likert items for brushing or flossing did not increase the
reported Cronbach’s values.
Using the total composite median scores for each participant, a Spearman’s
correlation coefficient of 0.039 with a significance value of 0.699 was found. This output
indicates a very weak positive correlation between the level of expressed oral health
literacy and the frequency at which Kentucky Appalachian adults utilize oral hygiene
techniques. This result is not statistically significant as indicated by p = 0.699. For
Research Question 1, the null hypothesis of ‘There is no relationship between a subject’s
expressed level of oral health literacy and the frequency in which they practice oral
hygiene techniques’ is accepted.
90
Research Question 2: Is there an association between the participant’s expressed
level of oral health literacy and the frequency at which their children practice oral
hygiene techniques?
Ha2: There is a relationship between the participant’s expressed level of oral
health literacy and the frequency that their children practice oral hygiene
techniques.
H02: There is no relationship between the participant’s expressed level of oral
health literacy and the frequency that their children practice oral hygiene
techniques.
The variable of expressed oral health literacy level was already computed as described
for Research Question 1. For the dependent variable of oral hygiene technique usage by
children, the following survey items were included:
My kids brush their teeth twice every day.
My kids floss their teeth every day.
Sometimes my kids don’t brush their teeth every day.
Sometimes my kids don’t floss their teeth every day.
A total composite median score was calculated for each participant that included the
responses from each of the four included survey items pertaining to oral hygiene methods
used by children. Participants who indicated that they did not have children were not
included in this scale. It is again of note that for the purposes of this study, the definition
of proper oral hygiene usage is considered to consist of brushing twice per day and
flossing once per day as recommended for proper results by the American Dental
91
Association (2016). Any differences in the reported rates between brushing and flossing
are covered later in this chapter. The Cronbach’s alpha scores for each included item are
as follows:
Table 3
Cronbach's alpha scores for Research Question 2
Perceived Risk Perceived as Preventable Perceived Self Control Child Brush/Floss
0.424 0.522 0.574 0.544
The omission of any of the Likert items for this scale did not increase the reported
Cronbach’s alpha values. Using the total composite median scores for each participant, a
Spearman’s correlation coefficient of .239 with a significance value of .017 was found.
By examining this output, it is determined that there is a weak, positive correlation
between the expressed level of oral health literacy and the frequency at which children
use oral hygiene techniques, rs = .239. This result is statistically significant as indicated
by p = .017. For Research Question 2, the null hypothesis of ‘There is no relationship
between the participant’s expressed level of oral health literacy and the frequency that
their children practice oral hygiene techniques’ is rejected.
Research Question 3: Are Kentucky Appalachian adults and children
practicing oral hygiene techniques at the same frequencies?
Ha3: There is a difference in how often oral hygiene techniques are used between
adults and children.
H03: There is no difference in how often oral hygiene techniques are used
between adults and children.
92
For Research Question 3, a Spearman’s correlation coefficient like what was used in the
first two research questions was not an appropriate method for analysis. To answer this
research question the frequencies of how often oral hygiene techniques were reported
being used were compared between participants and their children to determine if there
were differences between the two groups.
Table 4
Reported usage of oral hygiene techniques for adults and for children.
Reported adult oral hygiene usage rate at recommended frequencies.
N Percent Cronbach's alpha
Never 30 7.6%
.571
Rarely 76 19.2%
Sometimes 186 47.0%
Most of the time 81 20.4%
Always 23 5.8%
Reported child oral hygiene usage rate at recommended frequencies.
N Percent Cronbach's alpha
Never 91 32.6
.544
Rarely 76 27.2
Sometimes 62 22.2
Most of the time 43 15.4
Always 7 2.6
Here we see that 73.8% of adults report utilizing oral hygiene techniques at rates
of sometimes or less, while 82.0% of children are reported as utilizing oral hygiene
techniques at the same rates. However the distributions between the groups are different:
there is a higher percentage of children reported as utilizing techniques at rates of rarely
or never than adults. For Research Question 3, the null hypothesis is rejected.
Research Question 4: Are there differences in the survey responses based on
the participant’s county of origin, translational County A or distressed County B?
93
Ha4: There are differences in the survey responses collected between the
transitional and distressed counties.
H04: There are no differences in the survey responses collected between the
transitional and distressed counties.
Independent variable: County A and County B groupings.
Dependent variable: Responses to each survey item.
This last research question was developed to determine if any of the survey items were
answered differently based on where the survey was filled out, either County A or
County B. This would serve to demonstrate if answers varied between the transitional and
distressed counties. Here the Mann-Whitney U test was utilized as this test allows for the
comparison of differences between two independent groups when the dependent variable
is ordinal in nature (Laerd Statistics, 2018).
To use the Mann-Whitney U test, there are four assumptions about the data that must be
met (Laerd Statistics, 2018):
Assumption #1: There is one dependent variable that has been measured at the
ordinal level. This assumption has been met as this study has variables that have been
measured with Likert items which are ordinal in nature.
Assumption #2: There is an independent variable that consists of two categorical,
independent groups. This assumption is met in this study as the independent groups
consist of the responses gathered from County A and County B.
94
Assumption #3: There is an independence of observations between the two
independent groups. This assumption is met as there were different participants in both
county groups. No one participant was included in both groups.
Assumption #4: The distribution of scores of both independent groups needs to be
confirmed as being similar or different as this dictates how the results can be interpreted.
Using SPSS software, population pyramid charts were generated while calculating the
Mann-Whitney U tests. Each population pyramid chart was visually inspected to
determine if the two distributions contained in each Mann-Whitney U test were similar or
different (Laerd Statistics, 2018). For tests that demonstrated that the distributions were
similar, the medians were investigated (Laerd Statistics, 2018). For distributions that
were deemed visually different, the mean ranks were taken into consideration instead of
looking for differences in the median values (Laerd Statistics, 2018). In this study, U test
statistics could run from 0, indicating a complete separation between groups and that the
H0 should be rejected, and 2,438, indicating complete agreement between groups and that
the H0 should be accepted (Boston University, 2017). Overall the deciding factor for each
test regardless of distribution was the resulting significance value, p.
Using SPSS software, the Mann-Whitney U test was used to test the null
hypothesis for each item that stated there were no differences in the survey responses
collected in County A and County B. The result for each item is as follows with items
with different distributions listed with their mean rank scores:
95
Table 5 Mann-Whitney U test results of survey items based on county groups: Retain null
Survey
Item U Statistic Sig.
County A
mean rank
County B
mean rank Decision
2. I floss my teeth every day.
1209 0.94
Retain
4. Sometimes I am just too busy to brush my teeth twice a day.
1094 0.338
Retain
6. My kids floss their teeth every day.
1523 0.242
Retain
7. If I brush and floss every day, I will have good teeth.
1447 0.091
Retain
11. I have had cavities in my teeth.
1095 0.328
Retain
13. Sometimes I am just too busy to floss my teeth every day.
1373 0.238 47.08 53.36 Retain
18. As an adult I can still have poor teeth even if I brush and floss a lot.
1143 0.562
Retain
19. Kids can only have good teeth if they see a dentist a lot.
1230 0.934
Retain
23. Having poor teeth can make me sick in other ways.
1035 0.162
Retain
96
Table 6
Mann-Whitney U test results of survey items based on county groups: Reject null
Survey
Item U Statistic Sig.
County A
mean rank
County B
mean rank Decision
1. If parents have poor teeth, their children will too.
1035 0.004
Reject
3. Having cavities is a normal part of life.
807 0.002
Reject
5. Having poor teeth can lead to other health problems.
632 0.002 42.2 58.99 Reject
8. My kids have had cavities in their teeth.
449 0.025
Reject
9. Sometimes my kids don't brush their teeth twice a day.
534 0.016 38.57 33.21 Reject
10. People are born with either good teeth or bad teeth and they stay that way for their whole life.
952 0.011 56.56 42.45 Reject
12. I brush my teeth twice every day.
817 0.002
Reject
14. Sometimes having poor teeth just happens and people can't help that.
827 0.004
Reject
15. I can only have good teeth if I see a dentist a lot.
562 0.01 43.52 57.47 Reject
16. If kids brush their teeth every day while they are young, they will have better teeth as they age.
770 0 39.6 61.98 Reject
17. If kids floss their teeth every day while they are young, they will have better teeth as they age.
803 0 38.98 62.71 Reject
20. Sometimes my kids don't floss their teeth every day.
976 0 26.61 46.22 Reject
21. My kids brush their teeth twice every day.
875 0.003 29.35 43.24 Reject
22. If someone has bad teeth, they can only get better by seeing a dentist.
928 0.034 55.48 43.68 Reject
The significance of these results and how they identify and relate to overall trends are
discussed later in this chapter and in Chapter 5 of this study.
Descriptive Information
While this study was driven by four main research questions with corresponding
hypotheses, it also served to provide descriptive information into how Kentucky
Appalachians perceive several oral health related topics. This data further assists in filling
the gap in the currently available literature by providing insight into any trends Kentucky
Appalachians may display in how they feel about different aspects of oral health and how
it pertains to them. Descriptive data such as the following can assist policy-makers and
97
public health officials to better understand public perceptions and behavior by providing
insight into what issues to focus on when designing public health programs in local
communities (Rubin et al., 2014). The following data is comprised of discrete variables,
or variables that can only be specific values such as what was obtained on the survey
used in this study using Likert type items (Boston University, n.d.; Hussain, 2012). As
such, the data was analyzed and trends identified by examining the frequency and
percentages of the responses to the corresponding survey items (refer to Appendix for the
survey items used in this study).
Much like that was done for each research question, a Likert scale was developed
for each descriptive item to avoid depending on any one survey item to provide results.
Using SPSS software, the frequencies for each Likert scale were found. As an example,
for descriptive item 1 there were four Likert-type survey items used in the corresponding
Likert scale and the responses for each item were included in the frequency count to
provide an answer to the descriptive item. Any significantly different results between the
responses of both counties as found by the Mann-Whitney U test are also discussed. The
Cronbach’s alpha and IQR are also provided for each Likert scale. The results for each
descriptive item are as follows:
1. Do Kentucky Appalachian adults practice oral hygiene methods at
recommended rates?
The included survey items asking about oral hygiene rates were:
I floss my teeth every day.
I brush my teeth twice every day.
98
Sometimes I am just too busy to brush my teeth twice a day.
Sometimes I am just too busy to floss my teeth twice a day.
After reversing the scores of the last two items, the frequencies for each answer choice
were found and the results compared based on the county of origin:
Table 7
Descriptive Item 1 Results.
Do adults practice oral hygiene methods at recommended frequencies?
Response n Percent alpha IQR
Never 30 7.6 0.571 1
Rarely 76 19.2
Sometimes 186 47
Most of the time 81 20.4
Always 23 5.8
Significant differences found between the responses of County A and County B.
Mann-Whitney U test result
Survey Item U statistic Sig.
I brush my teeth twice every day.
817 0.002
From the data it is evident that 47.0% of the responses were the middle answer of
‘sometimes’. The remaining answers were nearly equally scattered among the remaining
answer choices with just a few more reporting an answer of ‘never’ than ‘always’. Here
we conclude that Kentucky Appalachian adults are likely not utilizing oral hygiene
techniques at recommended frequencies, indicating that they could possibly increase their
overall oral health status if frequencies were increased. A significant difference was
found in that County A was more likely to respond positively to the statement ‘I brush
my teeth twice every day’, however both counties had equally subpar responses to the
reverse statement.
99
2. Do Kentucky Appalachian children practice oral hygiene methods at
recommended rates?
Using the following survey items:
My kids brush their teeth twice every day.
My kids floss their teeth every day.
Sometimes my kids don’t brush their teeth every day.
Sometimes my kids don’t floss their teeth every day.
Table 8
Descriptive Item 2 Results.
Do children practice oral hygiene techniques at recommended frequencies?
Response n Percent alpha IQR
Never 91 32.6 0.544 1
Rarely 76 27.2
Sometimes 62 22.2
Most of the time 43 15.4
Always 7 2.6
Significant differences found between the responses of County A and County B.
Mann-Whitney U test result
Survey Item U statistic Sig.
My kids brush their teeth twice every day.
875 0.003
Sometimes my kids don't brush their teeth every day.
534 0.016
Sometimes my kids don't floss their teeth every day.
976 0
For this inquiry, no one answer can be pinpointed in the response frequencies as
being clearly chosen more often than the rest. However, it is found in that an
overwhelming 82.0% of the responses indicate that children are ‘sometimes’, ‘rarely’ or
‘never’ utilizing oral health techniques at recommended frequencies. Further
100
investigation into the differences between counties demonstrated significant results.
Participants in County B were more likely to respond positively to the statement ‘My kids
brush their teeth twice every day’, however they did not express the same trend when
answering the reverse statement ‘Sometimes my kids don’t brush their teeth every day’.
County A was more consistent in their answers to both statements. County B participants
were more likely to disagree with ‘Sometimes my kids don’t floss every day’, however
both counties answered fairly equally with subpar responses to ‘My kids floss every day’.
For this behavioral investigation, it was found that participants in County B were not
consistent in their responses. However this data shows that Kentucky Appalachian
children are likely not utilizing oral hygiene techniques at recommended frequencies.
3. Do Kentucky Appalachian’s perceive dental decay as a health risk?
The following survey items were analyzed:
Having poor teeth can lead to other health problems.
Having poor teeth can make me sick in other ways.
101
Table 9
Descriptive Item 3 Results.
Do Kentucky Appalachian's perceive dental decay as a health risk?
Response n Percent alpha IQR
Never 28 14.1 0.424 1.5
Rarely 55 27.8
Sometimes 41 20.7
Most of the time 48 24.2
Always 26 13.2
Significant differences found between the responses of County A and County B.
Mann-Whitney U test result
Survey Item U statistic Sig.
Having poor teeth can lead to other health problems.
632 0.002
A clear answer cannot be determined from the supplied responses. The trend here
is that responses are nearly equal across ‘strongly agree’ and ‘strongly disagree’ and
‘agree’ and ‘disagree’. This lack of one true response is also evident by the lower
Cronbach’s alpha value of .424 and a higher IQR of 1.5. This indicates that there is
nearly the same number of people that feel that dental decay is not a health risk as there
are people who do feel it is a health risk. However, when considering the Mann-Whitney
U test results, it was demonstrated that participants in County A were more likely to
respond positively to both included items. County B participants were more likely to
disagree with both statements, significantly so with ‘Having poor teeth can lead to other
health problems’.
4. Do Kentucky Appalachian adults feel that dental decay is preventable?
The following survey items were included:
If parents have poor teeth their kids will too.
102
If I brush and floss, I will have good teeth.
If kids floss their teeth while they are young, they will have better teeth as
adults.
If kids brush their teeth while they are young, they will have better teeth as
adults.
Table 10
Descriptive Item 4 Results.
Do Kentucky Appalachian's perceive dental decay as a health risk?
Response n Percent alpha IQR
Never 26 6.6 0.522 1
Rarely 77 19.4
Sometimes 104 26.3
Most of the time 149 37.6
Always 40 10.1
Significant differences found between the responses of County A and County B.
Mann-Whitney U test result
Survey Item U statistic Sig.
If parents have poor teeth their kids will too.
1035 0.004
If kids floss their teeth while they are young, they will have better teeth as adults.
803 0
If kids brush their teeth while they are young, they will have better teeth as adults.
770 0
The trend for this descriptive item is found by looking at the total of responses
that indicated agreement or disagreement. Here, 47.7% of the responses indicate
agreement that dental decay is preventable opposed to 26% who don’t feel that it is
preventable. Unfortunately, over a quarter of responses indicated that the participants
answered with ‘I don’t know’. The results are largely significantly different based on the
participant’s county of origin with more people in County A answering with ‘I don’t
103
know’ than County B where participants were more likely to agree to all included
statements.
5. Do Kentucky Appalachian adults feel that they can personally decrease the
risk of dental decay?
Included the following survey items:
I can only have good teeth if I see a dentist often.
Kids can only have good teeth if they see a dentist often.
Table 11
Descriptive Item 5 Results.
Do Kentucky Appalachian's feel they can personally decrease the risk of dental
decay?
Response n Percent alpha IQR
Never 3 1.5 0.574 1
Rarely 29 14.6
Sometimes 43 21.7
Most of the time 96 48.5
Always 27 13.6
Significant differences found between the responses of County A and County B.
Mann-Whitney U test result
Survey Item U statistic Sig.
I can only have good teeth if I see a dentist often.
562 0.01
The results for the concept of self-control indicate that Kentucky Appalachian adults are
more likely to feel that only by seeing a dentist can they decrease the risk of dental decay,
although almost 22% answered with ‘I don’t know’. However, it was found that people in
County B were more likely to feel this way over people in County A with their own
dental health. Both counties were similar in their disagreement that children can only
have good teeth if they see a dentist often.
104
6. Do Kentucky Appalachian adults perceive oral hygiene techniques as
important for good overall health?
To investigate this concept, the following Likert survey items were included:
If I brush and floss every day, I will have good teeth.
People are born with either good teeth or bad teeth and they stay that way for
life.
I can only have good teeth if I see a dentist a lot.
I can still have poor teeth even if I brush and floss often.
Having poor teeth can make me sick in other ways.
Table 12
Descriptive Item 6 Results.
Do Kentucky Appalachian's view oral hygiene as important for overall health?
Response n Percent alpha IQR
Never 29 5.9 0.518 1.5
Rarely 80 16.2
Sometimes 119 24
Most of the time 198 40
Always 69 13.9
Significant differences found between the responses of County A and County B.
Mann-Whitney U test result
Survey Item U statistic Sig.
People are born with either good teeth or bad and they stay that way for life.
952 0.011
I can only have good teeth if I see a dentist often.
562 0.01
It was found that 53.9% of the collected responses indicated that they agreed that oral
hygiene techniques are important for overall health. However, this left roughly 46.0% of
the participants as either not knowing or disagreeing. County A participants were
105
significantly more likely to respond with ‘I don’t know’ to the item ‘People are born with
either good teeth or bad teeth and they stay that way for life’, while again, County B
participants were more likely to indicate that they believe proper oral health can only be
achieved by seeing a dentist often.
7. To what level do Kentucky Appalachian adults perceive that enforcing
childhood oral hygiene techniques will decrease the risk of dental decay as
their children age?
Using the following Likert type survey items:
If kids floss their teeth while they are young, they will have better teeth as
adults.
If kids brush their teeth while they are young, they will have better teeth as
adults.
Table 13
Descriptive Item 7 Results.
Do adults perceive that proper childhood oral hygiene leads to better lifetime oral
health?
Response n Percent alpha IQR
Never 4 2 0.661 1
Rarely 24 12.1
Sometimes 51 25.8
Most of the time 96 45.5
Always 23 11.6
Significant differences found between the responses of County A and County B.
Mann-Whitney U test result
Survey Item U statistic Sig.
If kids floss their teeth while they are young, they will have better teeth as adults.
803 0
If kids brush their teeth while they are young, they will have better teeth as adults.
770 0
106
The responses indicate that overall, more people feel that oral health techniques
do help children to decrease their risk of dental decay as they age. However, roughly a
quarter of the participants included in this study indicated that they don’t know if these
techniques will help or not. Participants in County B were more likely to agree with both
statements while more County A participants responded with ‘I don’t know’.
8. Do Kentucky Appalachian adults perceive poor dental health as a normal
event?
The survey items included for this concept were:
Sometimes having poor teeth just happens and people can’t help that.
Having cavities are a normal part of life.
People are either born with good teeth or poor teeth and they stay that way for
life.
Upon calculation of the Cronbach’s alpha of .183, it was determined that these questions
have no correlation to each other and that they do not serve to measure the same factor.
The Cronbach’s alpha value is too low and the removal of any of the included items does
little to improve the value.
Summary
After analyzing the data for this study, many noteworthy trends emerged from the
research questions and supporting descriptive items. Descriptive Item 1 focused on the
frequencies that Kentucky Appalachian participants utilized oral hygiene techniques. It
was found that adults in both counties are largely not brushing or flossing at frequencies
recommended by the American Dental Association. The reported 73.8% of the responses
107
across both counties indicated that the majority of respondents utilized these techniques
only ‘sometimes’ or less often. By examining the Likert type items included in this scale
individually, it was found that the frequency of brushing were higher than frequencies of
flossing. Both counties reported equally subpar flossing habits. This indicates that if
frequencies of brushing and flossing could be increased, this may assist in reducing the
occurrence of poor dental health and disease found in people located in Appalachian
areas.
Descriptive Item 2 investigated how often Kentucky Appalachian children
utilized oral hygiene techniques. Rates of flossing were not statistically significantly
different across the counties. County A had more responses of disagreement to the
statement ‘Sometimes my kids don’t brush their teeth every day’ and this was found to be
significantly different from County B. However, by examining the ratio of provided
answers, 82.0% of the parental responses indicated that their children are utilizing these
techniques at frequencies of ‘sometimes’ or less often. By examining the Likert type
items individually, again it was found that rates of flossing were particularly subpar in
both counties. This data indicates that the usage of childhood oral hygiene techniques
could be improved upon and that much like for adults, this may help in decreasing the
rate of poor dental health and disease in children in these areas.
Descriptive Item 3 was formulated to investigate if Kentucky Appalachian adults
perceive dental decay as a health risk. The IQR value of 1.5 and a lower Cronbach’s
alpha value of .424 indicated that the responses were scattered instead of showing a clear
pattern. This was indeed true when looking at the response frequencies. When answering
108
the two statements of ‘Having poor teeth can lead to other health problems’ and ‘Having
poor teeth can make me sick in other ways’, 20.7% of the participants answered with ‘I
don’t know’. There were slightly more participants who responded with ‘disagree’ or
‘strongly disagree’ than people who ‘agreed’ or ‘strongly agreed’ to these statements.
However it was found that County B had more participants who agreed with these
statements, with the results for ‘Having poor teeth can lead to other health problems’
being statistically significantly different from County A. One reason for this response
pattern might be that people in County B may be more likely to have had firsthand
experience of the additional health issues that can result from poor teeth than people in
County A. This theory is supported by the fact that County B has almost 2.5 times the
rate of tooth loss, defined as the percentage of adults missing six or more teeth, than
County A (Kentucky Health Facts, 2018). Localized areas in County B report that over
50.0% of adults suffer from tooth loss.
Descriptive Item 4 investigated to determine if Appalachian adults felt that dental
decay is a preventable health risk. People in County B were more likely to agree to the
statements of ‘If kids floss their teeth while they are young, they will have better teeth as
adults’ and ‘If kids brush their teeth while they are young, they will have better teeth as
adults’ and these were statistically significantly different from County A who were more
likely to respond with ‘I don’t know’. County A was more likely to disagree with the
statement of ‘If parents have poor teeth their kids will too’ and this was statistically
significantly different from County B. By looking at the frequencies for the items
included in this Likert scale, it was found that 52.3% of respondents either didn’t know or
109
disagreed with the statement. This led to Descriptive Item 7 which looked at if Kentucky
Appalachian adults feel that enforcing childhood oral hygiene techniques leads to a
reduced risk of dental decay as their children age. Again, the answers to these included
items showed statistically significantly different answers between counties, with County
B more likely to respond positively while more people in County A responded with ‘I
don’t know’. Although County B was found to be more likely to respond positively to the
statements included in these two related research questions, their answers provided to the
items focusing on childhood oral hygiene usage rates indicate that their children actually
utilize these techniques at rates lower than County A. Descriptive Item 5 gives more
insight into this trend.
Descriptive Item 5 looked into if Kentucky Appalachian adults feel they can
personally decrease their risk of dental decay. To the item ‘I can only have good teeth if I
see a dentist a lot’, 78.2% of the respondents in County B were more likely to agree or
strongly agree and this result was statistically significantly different from County A. This
may assist in explaining the results of Descriptive Items 4 and 7 and the reported poor
correlation to oral hygiene frequencies: people in County B may be more likely to see
dental decay as a health risk but are more likely to believe that good teeth can only be
obtained by seeing a dentist regularly and so may be less likely to utilize oral hygiene
techniques at recommended frequencies.
Descriptive Item 6 investigated to see if Kentucky Appalachian adults viewed oral
hygiene techniques as being important to overall good health. After examining the
frequencies for all five survey items included for this question, it was again found that
110
County B was more likely to respond as good teeth can only be obtained by seeing a
dentist regularly. Overall, 53.9% of responses indicated that oral hygiene techniques are
important for good overall health. However, this does leave nearly half of the responses
indicating not knowing or disagreeing to this concept. This may be a factor behind the
high rate of non-utilization of oral hygiene techniques that was found in Research
Question 1. People in the included areas may not be brushing and flossing at rates as
recommended by the American Dental Association because nearly half of them don’t feel
that these techniques are important for overall good health.
As previously identified, Descriptive Item 7 sought to find out if Kentucky
Appalachian adults felt that enforcing childhood oral hygiene techniques was important.
Overall, it was found that 60.1% of the responses agreed or strongly agreed to this
concept. However, the vast majority of the responses of ‘I don’t know’ were collected in
County A, which was statistically significantly different from County B. County B was
more likely to respond positively to this concept, but again, this isn’t reflective in their
reported rates of oral hygiene techniques being utilized by their children.
Descriptive Item 8 sought to find out if Kentucky Appalachian adults perceived
poor dental health as a normal health event. This question could not be answered in its
entirety due to a very low Cronbach’s alpha value. All three of the included items that
made up this Likert scale were found to have the highest IQR values:
111
Table 14
Descriptive Item 8 IQR Percentiles per survey item.
4. Having cavities is a normal part of life.
Percentiles: 25 2.00
50 2.00
75 4.00
11. People are born with either good teeth or bad teeth and they stay that way for life.
Percentiles: 25 1.00
50 2.00
75 3.00
15. Sometimes having poor teeth just happen and people can't help that.
Percentiles: 25 2.00
50 3.00
75 4.00
This indicates that the responses were scattered across the answer choices and a trend is
difficult to pinpoint. This was found to be true when looking at the frequencies for each
item:
112
Table 15
Descriptive Item 8 response frequencies.
4. Having cavities is a normal part of
life.
n
Strongly disagree 10
Disagree 40
I don't know 23
Agree 22
Strongly agree 4
11. People are born with either good teeth or bad teeth and they stay that way for life.
n
Strongly disagree 25
Disagree 35
I don't know 28
Agree 9
Strongly agree 2
15. Sometimes having poor teeth just happen and people can't help that.
n
Strongly disagree 3
Disagree 34
I don't know 16
Agree 32
Strongly agree 14
The responses are nearly equal across agree and disagree, with many indicating that they
don’t know. The only trend from this data that can be identified is that there is no true
agreement among the items included in this scale.
The output produced for Research Question 1 suggested a very weak positive
correlation between the expressed level of oral health literacy and the frequency at which
Kentucky Appalachian adults utilize oral hygiene techniques, rs = .039. This result was
not found to be statistically significant as indicated by p = .699. For Research Question 1,
the H0 of ‘There is no relationship between a subjects expressed level of oral health
literacy and the frequency in which they practice oral hygiene techniques’ was retained.
113
In Research Question 2, it was determined that there was a weak, positive
correlation between the expressed level of oral health literacy and the frequency at which
children use oral hygiene techniques, rs = .239. This result was statistically significant as
indicated by p = .017. For Research Question 2, the H0 of ‘There is no relationship
between a subjects expressed level of oral health literacy and the frequency that their
children practice oral hygiene techniques’ was rejected. However, it was found that
children are utilizing oral hygiene techniques at frequencies that are much lower than
recommended. These results suggest that while a parent’s oral health literacy level may
be influencing how often their children brush and floss, it may be that the parent’s overall
oral health literacy level may be too low to promote proper oral hygiene technique usage
in their children.
The frequency tables that were calculated for Descriptive Items 1 and 2 were
utilized to answer Research Question 3. Upon examination of the distribution of
responses, it was determined that there was a difference between adults and children in
their utilization of oral hygiene techniques. For Research Question 3, the H0 of ‘There is
no difference in the usage rate of oral hygiene techniques between subjects and their
dependent children’ was rejected as more children are reported as not using oral hygiene
techniques appropriately. As presented in Chapter 1, it is documented that children
largely learn health habits and attitudes from their parents, particularly during their first
few years of life as they are completely dependent on their caregiver’s for all health
related treatments (Rhee, 2008). This documented trend can be seen in the data for
Research Question 3: children are not using oral hygiene methods in higher frequencies
114
than their parents and in fact, the data demonstrated that they are utilizing these methods
less often than their parents. This indicates that educational programs designed to assist in
improving the transition of dental related knowledge from parent to child may be of use
in Kentucky Appalachian communities. It is of note that was found that children have
higher rates of teeth brushing, particularly in County B, than flossing; however the
reported rates of childhood flossing in both counties were highly subpar to the reported
rates of flossing in adults.
Research Question 4 involved the comparison of each survey item based on
county A or B response origin. In this way it was determined whether there was a
statistically significant difference in responses based on county designation. One trend
that appeared during these calculations was that County A was much more likely to
respond with ‘I don’t know’. County B was more likely to have more definitive answers
of ‘agree’ or ‘disagree’ then to choose ‘I don’t know’ as a response. There were a few
interesting differences to note where the H0 of ‘There are no differences in the survey
responses collected between County A and County B’ was rejected.
For the item of ‘If parents have poor teeth, their kids will too’, County A was more likely
to disagree with this statement than County B. For the item of ‘Sometimes having poor
teeth just happens and people can’t help that’, County A was more likely to disagree than
County B. This suggests that County B may generally hold a more apathetic acceptance
of poor dental health than County A.
Conversely, County B demonstrated a better understanding of how childhood
utilization of oral hygiene techniques may decrease the occurrence of dental decay as
115
children age. For the statements of ‘If kids brush their teeth every day while they are
young, they will have better teeth as they age’ and ‘If kids floss their teeth every day
while they are young, they will have better teeth as they age’ County B was more likely
to agree to these concepts while County A had more ‘I don’t know’ responses. County B
also reported more positive responses to the item ‘My kids brush their teeth twice every
day’ than did County A. However, for the reverse item of ‘Sometimes my kids don’t
brush their teeth twice a day’, the trend did not hold for County B as there were more
County B responses of ‘sometimes’ and ‘most of the time’ for this item. County A
demonstrated more consistency between both items regarding childhood teeth brushing
habits.
Another trend that was spotted in the resulting data was that both counties
answered very similarly for the item ‘I have had cavities in my teeth’ with ‘sometimes’
being the most common answer. But for the item ‘My kids have had cavities in their
teeth’ the results were statistically significantly different in that County B reported much
higher responses of ‘rarely’ or ‘never’, while County A answered more with ‘sometimes’.
With a slightly lower dentist to patient ratio, a much higher percentage of adults suffering
from tooth loss, and the data by Dawkins et al. (2013) documenting that rural Kentucky
Appalachian children have an average of two untreated dental caries it is suspected that
this trend may be the result of not having seen a dentist to be properly diagnosed: parents
may be unaware of the true status of their children’s dental health. This theory cannot be
further investigated here as no data regarding dental professional visit history was
obtained in this study.
116
Overall, the answers from County A were more consistent, particularly with the
items that had matching reverse items, and with the items that had similar matching items
with different wording. It was investigated to see if a lack of education played into the
survey results, particularly in County B which has a lower rate of high school completion
than that of County A. However, 11 people from County A reported a high school degree
or less, while only three from County B reported the same. Chapter 5 will continue the
analysis, particularly in how this data relates to the Health Belief Model framework, how
it fits into the currently available literature and documentation, the limitations of this
study, and how it suggests the need for further investigation and research
117
Chapter 5: Final Interpretation
Introduction
Presented here is the final interpretation of the research study. The findings are
applied to the currently available documentation that was presented in the literature
review. Additionally, the limitations of this study are presented as well as
recommendations for future focus and research. Finally, the implications that this study
identified are discussed along with the potential impact the findings could have on the
included communities, as well as ways public health professionals could use these
findings to introduce positive social change.
Interpretation of the Findings
The results of this study served to confirm and extend the knowledge that was
previously documented and presented in the literature review contained in Chapter 2. A
concept that was investigated during the course of this study was the use of proper oral
hygiene methods in the included participant pool. Brushing twice per day and flossing
once per day are considered to be the most important and effective tools in the prevention
of dental decay and disease (Kidd, 2011). These prevention measures are essential to
begin when the first primary teeth appear in children, which can be as early as occurring
at six months of age, as both the primary teeth and the later appearing permanent teeth
erupt with immature enamel (Colgate-Palmolive, 2018; Kawashita, Kitamura, & Saito,
2011; Peterson-Sweeney & Stevens, 2010). This makes newly erupted teeth particularly
susceptible to bacteria and cavities until the enamel is adequately mineralized. Although
these oral hygiene methods are essential for proper oral health, there is little to no
118
information that details how often these methods are used, particularly in at-risk
populations, such as the Kentucky Appalachian population. Frisbee et al. (2010) and
Neiswanger et al. (2015) both evaluated for the use rates of oral hygiene methods in their
studies and documented the need to determine if deficiencies in usage exist. If
deficiencies are found, this provides public health officials an opportunity to advocate for
more frequent usage in order to decrease the risk of dental decay and disease. In this
study, it was evaluated for how often participants utilized oral hygiene methods. Of the
adult participants, 73.8% of the responses indicated that oral hygiene methods are being
utilized at frequencies of sometimes, rarely, or never. For the reported answers regarding
children’s oral hygiene habits, 82.0% of the responses indicated similar usage frequencies
as adults. The frequencies in children were statistically significantly different than adults
in that there were more answers of sometimes, rarely, and never. This data demonstrates
that Kentucky Appalachian children may not be using oral hygiene methods at the same
rates as their parents, let alone at recommended usage rates. The trend of poor usage rates
as reported in this study assists to confirm the accepted belief that attitudes towards oral
care are being passed down from parent to child and creating a cycle of poor dental health
that is proving difficult to intercept (Blanton & Ricardson, 2011; Dye et al., 2011). This
data also supports more recent statistics that show that the overall dental health status in
some Kentucky Appalachian areas is actually worsening despite efforts such as school-
based sealant and mobile dental programs (Blanton & Ricardson, 2011; Kentucky Youth
Advocates, 2016). The extension in knowledge regarding oral hygiene methods that this
study provided suggests that usage rates of brushing and flossing could be increased in
119
the Kentucky Appalachian population, both in adults and children, in an effort to
decrease the occurrence of poor dental health.
Another concept that was investigated by this study, which was suggested by the
available literature, was the level of perceived importance that Kentucky Appalachian
adults hold towards oral hygiene. Savage et al. (2014) documented in their study, which
took place at an Appalachian Kentucky college, that many participants reported that
people in their communities did not place importance on proper oral health. This
sentiment was echoed in a report by Kentucky Youth Advocates (2005) in which they
documented that dental health is not considered a community priority and that this
attitude is largely spread across Kentucky and not just isolated to the Appalachian areas.
The study presented here found that the included participants were approximately split
evenly between agreeing and disagreeing in believing that sometimes having poor teeth
just happens and people can’t change that. They were also almost evenly split between
believing that cavities were a normal part of life. The data collected on these concepts
support and expand on the previous findings of Savage et al. (2014) and the Kentucky
Youth Advocates (2005) as participants in this study did not express strong positive
trends in their beliefs of what constitutes a proper oral health status.
The major focus of this research study was the attempt to evaluate for the level of
oral health literacy that was expressed by the included Kentucky Appalachian
participants. The need for collecting this data was greatly supported by the currently
available documentation, as well as the gaps that were identified in what was available.
As presented in the literature review, it has been suggested that poor oral health literacy
120
may be acting as a barrier to proper dental health and that much like health literacy, oral
health literacy levels may decrease in areas that have a poorer socioeconomic status, such
as is found in Kentucky Appalachian areas (Centers for Disease Control and Prevention,
2011; Jones et al., 2007; Milfrom et al., 2004). Throughout the literature review, it was
discovered that there is a limited amount of information that directly targets the concept
of oral health literacy, however it is generally accepted that oral health literacy is
dependent upon factors that are similar to what contributes to health literacy. It is
documented that the Kentucky Appalachian population holds an overall lower level of
health literacy, however, to date there is no readily available data regarding the overall
oral health literacy level that exists in this area (Lee et al., 2012; Ludke et al., 2006). It is
this gap in the current literature that greatly drove the development of this study.
It is accepted that a person’s health literacy level is dependent upon the amount of
knowledge a person holds towards health topics, which includes the ability to recognize a
health risk (U.S. Department of Health and Human Services, 2018). Therefore, to
evaluate for oral health literacy in this study, several concepts were taken into
consideration in order to evaluate for the knowledge level towards oral health in the
included participants. The first concept was if the included participants felt that dental
decay was a health risk that could lead to further negative health events. A clear trend
could not be determined from the provided responses: roughly the same number of
people who agreed with dental decay as being an overall health risk that could lead to
further health events also disagreed with this concept. This data suggested that there is a
121
deficiency in the ability to identify dental decay as a health risk that could pose additional
health events in the included Kentucky Appalachian population.
A second concept that was considered when evaluating for oral health literacy
level was if Appalachian adults felt that dental decay was a preventable health risk. Here,
the responses were nearly split equally between agreement and ‘I don’t
know’/disagreement. Again, the only trend that can be inferred from this data is that there
isn’t a clear understanding that dental decay can be preventable.
The third concept that was included when evaluating for oral health literacy was if
Kentucky Appalachian adults felt that they could personally decrease the risk of dental
decay or if they felt that their dental health status was dependent upon a dentist. Roughly
60% of the responses indicated the belief that their dental health status was dependent
upon a dentist, with another 22% indicating that they didn’t know. This inability to
recognize their personal role in proper oral health served to further describe the amount
of knowledge the included participant pool held towards oral health.
Although there were no similar studies to be found in order to compare results to,
after considering these three concepts it was concluded that the overall expressed oral
health literacy level of the included Kentucky Appalachian participants can be considered
as low. This conclusion was reached due to the lack of trends showing a strong positive
understanding of the included concepts. Additionally, the poor usage rates of oral hygiene
techniques was considered in this conclusion as the ability to engage in self-care health
related habits is accepted as being affected by a person’s health literacy level (U.S.
Department of Health and Human Services, 2018). Here, it is concluded that the poorer
122
understanding of the included concepts that was expressed by the included Kentucky
Appalachian participants is contributing to the poor usage rates of oral health techniques
that were reported. This study found that the oral health literacy level of the included
parents did have a statistically significant correlation to how often their children brushed
and flossed as indicated by p = .017. While there was no statistically significant
correlation found in the adult participants, this study documented that they are generally
not using oral hygiene methods in recommended frequencies. No statistically significant
differences could be identified in the expressed oral health literacy level based on county
designation. The expressed levels were generally similar in both the transitional and
distressed Appalachian counties.
Another trend of note that was discovered in this study’s data was that the
included Kentucky Appalachian participants generally expressed awareness regarding the
overall benefit of using oral hygiene methods both in themselves and their children.
However, these expressed views were not evident in the very poor reported usage rates of
oral hygiene methods. To assist in understanding this seemingly contradicting trend in the
data, the health belief model (HBM) theoretical framework was utilized in data
examination. To do this, the study presented here evaluated for the perceived self-
efficacy towards using oral hygiene methods that was present in the included participant
pool. The HBM’s construct of self-efficacy refers to people’s beliefs and confidence
regarding their own capability to obtain proper levels of performance or results that may
influence health events that could impact their lives (Bandura, 1994). As documented in
the literature review, people are more likely to adopt positive health habits if they have a
123
higher level of self-efficacy regarding the health habit (Glanz et al., 2008; Schunk &
Pajares, 2009). During the course of this study, it was found that roughly 62% of the
included Appalachian adults felt that only by seeing a dentist frequently could they obtain
proper dental health. Distressed County B respondents were more likely to feel this way
than respondents from transitional County A and these results were significantly different
from each other. This expressed correlation between dentists and proper oral health status
suggested that a lower level of self-efficacy towards utilizing oral hygiene methods
properly may exist. As an expansion on this suggestion, this study further investigated for
how Kentucky Appalachian parents felt that enforcing childhood oral hygiene methods
would correlate to children’s risk of dental decay. Roughly 60% of the responses
indicated agreement that enforcing childhood oral hygiene methods would benefit to
decrease the risk of dental decay as their children age. However, this is not reflected in
the reported usage rates of oral hygiene methods in children: the usage rates of oral
hygiene methods of children were lower than the reported usage rates of adults. This data
indicates that while many Kentucky Appalachian parents feel that oral hygiene methods
may benefit their children positively, they hold poor self-efficacy in their own ability to
enforce these habits in such a way that would properly improve their children’s oral
health. Additionally, to the survey item that stated ‘I can still have poor teeth even if I
brush and floss a lot’, roughly 69.0% of the responses indicated agreement. This suggests
that the included Appalachian adults also hold poor self-efficacy towards their ability to
conduct oral hygiene methods well enough to obtain the correct results in themselves. As
previously discussed, Glanz et al. (2008) and Schunk & Pajares (2009) maintain that
124
there is a direct correlation between a person’s self-efficacy and their likelihood to adopt
positive health habits. Based on this documentation and the data collected by the
presented study, it is predicted that Appalachian adults are currently unlikely to adopt
proper oral hygiene methods in themselves and their children due to a low level of
perceived self-efficacy towards these methods.
On top of contributing to the poor usage rates of oral hygiene techniques, the
discovery of a low level of self-efficacy also supports the findings of the included
population having a subpar oral health literacy level. This is because a person’s self-
efficacy has been documented to correlate to their health literacy level as well as their
health habits and self-care behaviors (Reisi et al, 2016; Xu & Leung, 2018). Those with a
higher level of health literacy are shown to be more confident in performing positive
health related habits and are more likely to practice these self-care behaviors (Bohanny et
at., 2013; Campbell, Beardsley, Shaya, & Pradel, 2015; Reisi et al., 2016). As such, the
overall low self-efficacy that was expressed by the participants included in the presented
study serves to support the conclusion that Kentucky Appalachians hold a low level of
oral health literacy. While this study was designed to evaluate self-efficacy towards oral
hygiene techniques, it was not expected that the data would show such poor beliefs in
these self-care behaviors. This is a significant finding with further implications that are
discussed later in this chapter.
The study presented here also investigated for any statistically significant
differences in the survey responses between the two included counties. It is documented
that health literacy levels decrease in areas that are more socioeconomically distressed
125
(Horowitz & Kleinman, 2008; Milfrom et al., 2004). As County B holds an overall lower
socioeconomic level it was expected that differences in the answers between the two
counties would be due to County A having a higher socioeconomic status as well as a
higher general level of formal education among its residents, as indicated by its
designation of transitional. Thus, County A participants were expected to demonstrate an
overall higher level of oral health literacy when compared to their counterparts from
County B. While this study found few notable differences to this expectation that are
discussed later as limitations, the overall trend found in the data here was the lack of
differences in the responses between the two counties. This indicates that both counties
largely share the same opinions and attitudes towards the concepts that were investigated
for in this study. These findings demonstrate that County A and County B participants
had similar levels of oral health literacy and understanding of the related concepts
included in this study. At the beginning of this research study, there were four Kentucky
Appalachian counties that had reached the economic status of transitional. At the time of
this writing, there are now two counties in the Kentucky Appalachian regions that are
considered transitional: the other two have now fallen into the at-risk designation
(Appalachian Regional Commission, 2018b). No Kentucky Appalachian county is
designated as holding the highest two ranks of competitive or attainment. These findings
suggest that the level of oral health literacy may be largely the same across the entire
Kentucky Appalachian region, regardless of county rank designation. This indicates that
public health efforts geared towards proper oral hygiene are needed across the region and
126
should not just focus on the most distressed areas as the currently available literature
suggests.
Limitations of the Study
There are limitations to this study that were identified in the execution of this
research project. The first of these limitations is that this study may have been at risk of
social desirability bias. Social desirability bias is considered to be a respondent-related
source of bias that can occur in research, particularly in studies that include the collection
of self-reported participant responses to questions or statements (Grimm, 2010). It refers
to the tendency of participants to respond in a way that they may feel is more socially
acceptable as opposed to responding to how they actually feel towards a question or
concept. Participants may answer differently if they feel they may be viewed negatively
were they to express their true feelings or beliefs. The reason it is suspected that social
desirability bias may exist in the presented study is that certain items contained in the
survey were not answered in such a way that was consistent with documented trends that
were reported by official agencies. As example, to the survey item ‘My kids have had
cavities in their teeth’, the responses were statistically significantly different, p = .025, in
that 68% of the responses from County B indicated an answer of rarely or never,
compared to 38% from County A. However, it is documented that County B has over
double the amount of adult tooth loss, which is defined as the loss of six or more teeth,
than that of County A and that pockets of over three times the number of County A exist
(Kentucky Health Facts, 2018). Although the survey item used in this study inquired
about cavities and the state collected data reports tooth loss, the results of each do not
127
demonstrate a correlation to each other: an area experiencing such elevated levels of
tooth loss in adults likely should have elevated rates of cavities and tooth decay in
children. The data found here also does not support the results of the research by
Dawkins et al. (2013) who found that an average of two untreated dental cavities existed
in Kentucky Appalachian children, with the average increasing in the more rural
Kentucky areas, of which County B is representative. It could be that respondents from
County B felt the more socially acceptable response regarding their children and cavities
was to report rarely or never. The discrepancies found in the collected data could also be
a result of the Hawthorne effect, which describes how the knowledge of being in a
research study may influence a participant into answering differently than they would
outside of a research setting (McCambridge, Witton, & Elbourne, 2014). Knowing that
they were participating in research may have influenced the included participants into
answering how they thought they should answer as opposed to answering in a way that
would more accurately portray their opinions or feelings. This may be particularly true of
the participants from County B as they were largely more vocal in expressing that they
had never been asked to participate in a research study before.
This trend could also be a result of self-report bias in that the status of ‘have had
cavities’ or ‘never had cavities’ was determined by self-reported data. With a slightly
lower dentist to population ratio than County A, the participants in County B may not
regularly consult a dentist regarding their children’s oral health status and it is possible
that their children may have cavities and the parents not know it. To avoid this bias,
dental health records could have been used to determine dental decay history for each
128
participant. However, this process would have eliminated any participants if they had not
visited a dentist and thus had no documented dental records. Therefore the possibility for
social desirability bias or self-report bias to exist in this study is acknowledged,
particularly as another study that asked for the same information in a similar manner and
population could not be located for result comparison.
Another limitation to this study was that the polling areas were limited to two
counties. The results of this study would be more generalizable to the remaining
Kentucky Appalachian population if more locations were included. Also of note, during
the course of this study it was discovered that the responses were largely similar across
both of the included transitional and distressed county. In light of this trend, it would
have been beneficial to poll in a non-Kentucky Appalachian location in order to provide
additional results to compare the Appalachian data to. This would have assisted in
demonstrating if the Appalachian responses truly were different. For example, by
comparing to a non-Appalachian location, it could have been more clearly determined if
the seemingly low oral health literacy level that was expressed by the included
Appalachian participants is actually low, or if it was similar to the rest of the state.
A final limitation to this study may be found in the instrument used for data
collection. As previously demonstrated throughout this study, a pre-existing instrument
designed to investigate for oral health literacy and the related concepts included in this
study could not be readily found and so a survey was designed specifically for use here.
Although much care was used when developing the survey items to attempt to ensure
reliability among the survey items all while maintaining an easy to read format as the
129
Kentucky Appalachian region has a documented existence of low literacy levels, the
survey may be at risk of lower reliability than an established instrument. Having no
similar study to use as a comparison also introduced difficulties when deciding how best
to analyze and present the Likert scale items. In the end, composite mean scores and the
ratio of responses were largely used to display and report the trends found in the data.
This study served to expand upon established concepts while discovering new trends in
the targeted population. It is hoped that this study will be of value for anyone who desires
to further research and expand upon the included concepts and findings in this
population.
Recommendations
After conducting the presented study, several recommendations have been
identified that should be considered in future research and public health efforts to further
extend upon the gathered knowledge and trends that were documented here. The first
recommendation would be to further investigate why there was a trend of such poor self-
efficacy regarding the use of oral health techniques. As previously reported, this study
found that the included Kentucky Appalachian participants generally indicated awareness
regarding the benefits of using oral hygiene techniques but at the same time indicated
very poor usage rates of said techniques. One of the major findings of this research was
that Kentucky Appalachians are largely not brushing or flossing in frequencies that are
considered beneficial and that these habits are being passed on to children, indicating a
cycle of poor oral health. An investigation into the reasons behind the poor self-efficacy
Kentucky Appalachian adults are holding towards brushing and flossing would be
130
beneficial as this is effectively acting as an additional barrier between Kentucky
Appalachians and a more positive oral health status.
In the Behavioral Health Continuum of Care Model, both prevention methods and
treatment methods are listed as key elements to a person’s overall wellbeing and health
(Wyoming Department of Health, 2018). Prevention would include any intervention that
is delivered prior to the onset of a health event in an effort to prevent or decrease the risk
of a health event occurring. These prevention efforts are largely seen as the domain of
public health officials. Treatment is any intervention that is delivered after the onset of a
health event, and falls into the domain of medical doctors and personnel. The results of
this study showed that people in Kentucky County B were more likely to respond that
only by seeing a dentist could they obtain a good oral health status. However, what is not
clear is how they are viewing dental professional care; do they see it as preventative, such
as they can only have good teeth if they see a dentist regularly, or do they see it as they
can only have good teeth if treatment by a dentist is obtained after tooth decay has
occurred and when the tooth needs filling/removing. Knowing that Kentucky County B
has such an elevated rate of adult tooth loss, it is suspected that residents in this county
are more apt to view dentists similarly to medical doctors in that they view dentists as
after the health event has occurred treatment providers. It would be beneficial to obtain
more information into how Kentucky Appalachian’s view dental professionals as if it was
found that dentists are considered more as treatment providers it may help these
communities if effort was made to bring the realization that dentists can also provide
prevention methods to be used before dental decay occurs. This recommendation is
131
expanded to include an investigation into how dentists in the Kentucky Appalachian
region are marketing themselves. If it were found that dentists are largely providing
treatment after dental decay has set in, it may be beneficial to work with local dentists in
order to help present to their communities that dental care can be preventative in nature
instead of a treatment after the fact. While some areas of the Kentucky Appalachian
region may be lacking in dental professional coverage, in areas where they are present
they should be assisted and educated in how they may best introduce and build awareness
towards proper oral health and hygiene habits in their local communities.
An additional recommendation for future efforts is grounded in the limitation
portion of this chapter and that is to include more areas and polling locations. Including
more Kentucky Appalachian areas would provide more generalizability of the results to
the remaining Appalachian population. Polling from non-Kentucky Appalachian areas
would possibly provide better comparison data to identify stronger trends and issues as it
was found in this study that the results from a transitional and distressed county were
oftentimes largely similar to each other. Including more areas would greatly assist in
further grounding the suggestion that oral health literacy levels are similar across the
Kentucky Appalachian region that was established by this study.
Another recommendation to be employed in future research efforts would be to
physically inspect the oral health status of the included participants. While this
recommendation would require much more extensive efforts from qualifying
professionals, this information would be more accurate in establishing the true oral health
status of the included participants. If this method couldn’t be utilized, possibly due to
132
cost, time, or other limited resources, it would be beneficial to obtain past dental health
records of included participants. These could be used to establish the existence of past or
current dental disease. Either of these methods would assist in decreasing or eliminating
the risk of social desirability or self-reporting bias that may be present in this study as the
data would be collected from methods other than participant self-response. If these
methods had been utilized in this study, it may have been found that the participants from
County B actually do have an oral health status, including cavities past or present, in rates
that are reported by official agencies instead of the self-reported rates of a much better
oral health status that were collected here.
Further investigation as to what the perceived barrier(s) is(are) that is keeping
Kentucky Appalachians from utilizing oral hygiene techniques at recommended rates
would provide public health officials valuable information as to how to combat this issue.
Examples of possible perceived barriers could be lack of time to perform oral health
techniques, lack of tools such as toothbrushes, a lack of knowledge of how to perform
oral health techniques properly, or any combination of these and more. Research that
would narrow down on what perceived barriers exist would help public health officials to
better tailor educational efforts in this population.
A final recommendation that has come about as a result of this study has to do
with the methods of future research and investigation efforts should utilize. The study
presented here did not collect any qualitative type data: all survey items were designed
with close-ended response choices. However, there was a noteworthy observation
regarding the participant pool that was made during the data collection process. This
133
observation was in the general behavior differences between the participants of County A
and of County B. The participants in County A largely completed their surveys and
returned them mostly in silence: there was little to no additional conversation past the
invitation to take the survey and a brief instructional dialogue made by the researcher.
This behavior was found to be in stark contrast to the behavior of their counterparts in
County B. While the exact proportion was not tracked, it was the experience of the
researcher that almost every participant in County B desired for and struck up additional
conversation, which largely centered on stories of the participant/participant’s friend or
family member and their personal experiences with oral health issues. While neither
group expressed much interest into why they were being asked to take a survey, County
B participants as a whole expressed more positive feelings regarding being asked to be
included. It was quite evident to the researcher that the participants in County B were
eager to share their opinions, both by filling out the survey but especially by additional
conversation. In light of this trend, it is recommended that future investigators consider
using a face-to-face, qualitative method when gathering data in more rural areas of the
Kentucky Appalachian region. Based on the how participants in County B were so quick
to engage in conversation and volunteer information, it is believed that valuable, accurate
insight into concepts of interest may be obtained by simply asking. This method may not
be as successful in less rural areas of the Kentucky Appalachian region as participants
from County A were more content to fill out their survey and generally did not attempt to
engage in extra conversation or volunteer any additional information or opinions.
Additionally, this qualitative approach may be of use by public health officials when
134
designing educational outreach programs into more rural Kentucky Appalachian areas.
These efforts may be met with better results if they are delivered in a more personal,
face-to-face format that encourages verbal engagement with community members as
opposed to providing information by way of a leaflet, poster, or some other more
impersonal method. Regardless of the method used, it is strongly recommended that more
research is done to further investigate oral health literacy and how it is effecting the
Kentucky Appalachian population.
Implications
The results of this study demonstrated that the included Kentucky Appalachian
participants are utilizing oral hygiene methods at usage rates that are far below the
recommended frequencies. This trend was found to be present in both of the Kentucky
Appalachian counties that were included in this study and were not significantly different
based on their county rank designations of transitional and distressed. This information
suggests that attempts to increase the usage rates of oral hygiene methods could greatly
assist in decreasing the high rates of dental decay and disease that is occurring in the
included Kentucky Appalachian locations, as well as possibly being beneficial to the
remaining Kentucky Appalachian population. However, this study also demonstrated that
programs that simply inform people in the included Kentucky Appalachian locations that
they should brush twice a day and floss once per day likely will be met with lackluster
results. As this study documented, the included participants demonstrated that they did
indeed know that oral hygiene techniques should be used at recommended rates. The
reason for this glaring contradiction in data is unknown, as discussed it could be
135
attributed to social desirability bias, self-report bias, or it could simply be that the
participants had gleaned this information from some other source, such as a toothpaste
commercial or a poster they have seen in a doctor’s office. This study documented that
when it came to utilizing oral health techniques at recommended rates, it was found that
the included Kentucky Appalachian participants demonstrated poor self-efficacy towards
these prevention methods.
Combining this trend with the collected data that showed that the included
participants demonstrated a low understanding of concepts that are considered part of a
person’s oral health literacy level, it is concluded that there is an opportunity for public
health officials and professionals to assist in breaking the poor oral health cycle that is
affecting the included Kentucky Appalachian communities. The HBM dictates that
people are more likely to adopt positive health habits if they understand the connection
between the health habits and the health event and that they believe by utilizing the health
habits they will decrease their personal risk of the health event occurring. Applying the
HBM to this study, it is predicted that the included Kentucky Appalachian participants
are currently unlikely to adopt appropriate oral hygiene techniques until public health
officials and professionals:
1. Develop community based education programs that serve to bring awareness
to the fact that dental decay is a largely preventable health risk.
2. Along with this awareness, the programs also need to present the correlation
between oral hygiene techniques and dental decay: clean teeth are healthier,
longer lasting teeth.
136
3. Introduce and promote personal empowerment in the oral health status of
Kentucky Appalachian adults and their children. Stressing the control people
have over their own oral health status could assist in improving the poor self-
efficacy that was reported in this study.
The results of this study demonstrate that these are the main points that need to be
addressed and considered when developing public health dental education efforts in
Kentucky Appalachian communities. These efforts are needed to improve the poor usage
rates of oral hygiene habits that were reported in this study. While there are many factors
such as a lack of available dentists or poor diet that may contribute to poor oral health,
using oral hygiene methods is arguably the best defense a person has against this health
risk. As this study documented, the included Kentucky Appalachian participants reported
a very poor usage rate in oral hygiene techniques and so it stands to reason that a usage
rate increase in this population would lead to a decrease in the risk of poor oral health. To
decrease the risk of poor oral health in this population would introduce a very positive
social change. Individuals would benefit from this social change by possibly not
experiencing the pain, discomfort, and malnutrition that is associated with severe dental
decay and tooth loss. On a societal level, the cycle of poor oral health that has plagued
the Kentucky Appalachian population for generations may finally be cracked, if not
broken. While this change would not happen quickly, the wheel that is community
education to improve the self-efficacy and oral hygiene method usage rates in the
Kentucky Appalachian population needs to be set in motion. To allow it to remain
stagnant means more generations of Kentucky Appalachian people will suffer from
137
dental decay, disease, and tooth loss, as this study documented that poor oral hygiene
habits, and possibly the corresponding poor oral health literacy level, are being passed on
in the current generation of Kentucky Appalachian parents and children.
To develop and introduce new community programs would be no small feat: they
require manpower and resources that may not be readily available in abundance, if at all.
To that end, facilities that are already established in the Kentucky Appalachian locations
that were included in this study could possibly be utilized. Both of the Kentucky
Appalachian counties that were included in this study had a local health department,
neither of which currently have any sort of dental health awareness/educational program
in place. Although the Kentucky Appalachian County B had a slightly less dentist to
population ratio than County A, this decrease was not significantly different, meaning
that both counties have nearly the same dental professional coverage. Public health
professionals should work with the dentists that are presently practicing in these counties
to enlist their efforts to better present dental decay as a preventative disease, as well as
the correlation with proper oral hygiene techniques, to the surrounding communities. This
effort could introduce positive organizational social change to the included communities
as currently practicing dental health professionals may not have the knowledge or tools
needed to provide effective dental decay prevention education to their patients. Although
it was not a focus of this research, it is also worth noting that during the course of this
study it was discovered that neither of the Kentucky Appalachian counties that were
included have school-based, free, or income-based sealant programs in place. While
sealants are not a long-term solution to the dental decay epidemic that is occurring in
138
Kentucky Appalachian areas, the addition of programs such as these could provide local
dentists and public health professionals an additional opportunity to implement education
to their respective communities. It is efforts like these described that are needed in order
to decrease the rates of dental decay in the included Kentucky Appalachian areas and
thereby bring about positive social change, both on the individual and societal levels.
Conclusion
There are many documented and accepted reasons why the Kentucky Appalachian
population continues to suffer from elevated rates of dental decay and disease. These
reasons include poor insurance coverage, a lack of practicing dental professionals, and
the lower education and socioeconomic levels that are generally found in Kentucky’s
Appalachian region. These issues are difficult to overcome without extreme intervention
from the state and federal levels. This study sought, and found, additional factors that are
contributing to the poor dental health status of Kentucky Appalachian areas:
1. Brushing and flossing are the best methods of self-defense against dental decay
and the Kentucky Appalachian participants that were included in this study are
not utilizing these methods in sufficient rates.
2. This study also demonstrated that the included participants overall had a poor
level of oral health literacy and poor self-efficacy in the use of oral hygiene
methods.
Until the oral health literacy level is improved in Kentucky Appalachian communities,
the very high rates of dental decay and disease that are found in these areas are likely not
going to decrease. A starting point would be to promote proper oral hygiene techniques,
139
as well as focus on improving the demonstrated poor efficacy people in these areas hold
towards these methods. The promotion of dental decay as being largely self-preventable
instead of having to be treated after it has occurred may also benefit Kentucky
Appalachian communities. These are factors that public health professionals who are
located in Kentucky Appalachian areas can realistically target and seek to improve upon
in order to effectively begin to assist in decreasing the rate of dental decay and disease
that has for so long plagued their local communities.
140
References
Allen, I.E., & Seaman, C.A. (2007, July). Likert scales and data analyses. Quality
Progress. [Electronic source]. Retrieved from http://rube.asq.org/quality-
progress/2007/07/statistics/likert-scales-and-data-analyses.html
American College Health Association. (2016). Ecological model. Characteristics of a
Healthy Campus Initiative. [Electronic format]. Retrieved from
https://www.acha.org/HealthyCampus/Implement/Ecological_Model/HealthyCam
pus/Ecological_Model.aspx?hkey=f5defc87-662e-4373-8402-baf78d569c78
American Dental Association. (2016). Mouth healthy. Life Stages.[Electronic source].
Retrieved from http://www.mouthhealthy.org
Andale, A. (2016, February 21). Ecological validity: Definition and examples. Statistics
How To: Design of Experiments. [Electronic source]. Retrieved from
http://www.statisticshowto.com/ecological-validity/
Appalachian Regional Commission. (2014). Poverty rates in Appalachia, 2010-2014.
Research, Maps, and Data. [Electronic source]. Retrieved from
https://www.arc.gov/research/MapsofAppalachia.asp?MAP_ID=121
Appalachian Regional Commission. (2015). County economic status and number of
distressed areas in Appalachian Kentucky, fiscal year 2015. [PDF format].
Retrieved from
https://kydlgweb.ky.gov/Documents/ARC/CountyEconomicStatusandDistressAre
asFY2016Kentucky.pdf
141
Appalachian Regional Commission. (2016). The Appalachian region. Appalachian
Region. [Electronic source]. Retrieved from
https://www.arc.gov/appalachian_region/theappalachianregion.asp
Appalachian Regional Commission. (2018). County economic status, fiscal year 2018:
Appalachian Kentucky. Region Reports. [Electronic source]. Retrieved from
https://www.arc.gov/reports/region_report.asp?FIPS=21999&REPORT_ID=71
Appalachian Regional Commission. (2018b, June). Economic status, by county, fiscal
year 2019. County Economic Status and Distressed Areas in Appalachia. [PDF
format]. Retrieved from
https://www.arc.gov/images/appregion/economic_statusFY2019/CountyEconomi
cStatusandDistressAreasFY2019Kentucky.pdf
Baltimore County Public Schools. (2017, June). Key elements of a research proposal:
Quantitative design. Types of Quantitative Design. [Electronic source]. Retrieved
from https://www.bcps.org/offices/lis/researchcourse/develop_quantitative.html
Bandura, A. (1994). Self-efficacy. In Ramachaudran, V.S. (Ed.), Encyclopedia of human
behavior (Vol. 4, pp. 71-81). New York, NY: Academic Press.
Bartlett, J.E., Kotrlik, J.W., & Higgins, C.C. (2001, Spring). Organizational research:
Determining appropriate sample size in survey research. Information Technology,
Learning, and Performance Journal, 19(1), 43-50. [Electronic format]. Retrieved
from:
http://ezproxy.uky.edu/login?url=http://search.proquest.com.ezproxy.uky.edu/doc
view/219816871?accountid=11836
142
Berkman, N.D., Sheridan, S.L., Donahue, K.E., Halpern, D.J., & Crotty, K. (2011, July
19). Low health literacy and health outcomes: An updated systematic review.
Annals of Internal Medicine, 155(2), 97-107. doi: 10.7326/0003-4819-155-2-
201107190-00005
Blanton, J., & Ricardson, K. (2011, September 8). UK partners with others to train more
dentists in rural areas. University of Kentucky News. [Electronic source].
Retrieved from https://uknow.uky.edu/campus-news/uk-partners-others-train-
more-dentists-rural-areas
Bohanny, W., Wu, S.F., Liu, C.Y., Yeh, S.H., Tsay, S.L., & Wang, T.J. (2013,
September). Health literacy, self-efficacy, and self-care behaviors in patients with
type 2 diabetes mellitus. Journal of the American Association of Nurse
Practitioners, 25(9), 495-502. doi: 10.1111/1745-7599.12017
Borak, J., Salipante-Zaidel,C., Slade, M.D., & Fields, C.A. (2012, February). Mortality
disparities in Appalachia: Reassessment of major risk factors. Journal of
Occupational and Environmental Medicine, 54(2), 146-156. doi:
10.1097/JOM.0b013e318246f395
Boston University. (n.d.). Analyzing and presenting results from descriptive studies.
School of Public Health. [Electronic format]. Retrieved from
http://sphweb.bumc.bu.edu/otlt/MPH-Modules/QuantCore/PH717_Analysis-
Presentation-Descriptive/PH717_Analysis-Presentation-Descriptive_print.html
Boston University. (2017, May 4). Mann Whitney U test (Wilcoxon rank sum test).
School of Public Health. [Electronic format]. Retrieved from
143
http://sphweb.bumc.bu.edu/otlt/MPH-
Modules/BS/BS704_Nonparametric/BS704_Nonparametric4.html
Campbell, M.L., Beardsley, R.S., Shaya, F.T., & Pradel, F.G. (2015, May). Self-efficacy,
health literacy and adherence to self-care behaviors in kidney transplant
recipients. The Journal of the International Society for Pharmacoeconomics and
Outcomes Research, 18(3), 188-197. doi: 10.1016/j.jval.2015.03.1090
Carifio, J., & Perla, R. (2008, November 19). Resolving the 50-year debate around using
and misusing Likert scales. Medical Education, 42(12), 1150-1152. doi:
https://doi.org/10.1111/j.1365-2923.2008.03172.x
Carpenter, C.J. (2010, December). A meta-analysis of the effectiveness of health belief
model variables in predicting behavior. Health Communication, 25(8), 661-669.
doi: 10.1080/10410236.2010.521906
Center for Innovation in Research and Teaching. (2017). Key issues in quantitative
research. Development Resources. [Electronic source]. Retrieved from
https://cirt.gcu.edu/research/developmentresources/research_ready/quantresearch/
keyissues
Centers for Disease Control and Prevention. (2011, October). Oral health. Healthy People
2010 Data. [Electronic source]. Retrieved from
https://www.cdc.gov/nchs/healthy_people/hp2010/data2010.htm
Centers for Disease Control and Prevention. (2015, March 25). The social-ecological
model: A framework for prevention. The Social-Ecological Model. [Electronic
144
format]. Retrieved from
https://www.cdc.gov/violenceprevention/overview/social-ecologicalmodel.html
Chang, B.L., Bakken, S., Brown, S.S., Houston, T.K., Kreps, G.L., Kukafka, R., . , &
Stavri, P.Z. (2004, November 1). Bridging the digital divide: Reaching vulnerable
populations. Journal of the American Medical Informatics Association, 11(6),
448-457. doi: https://doi.org/10.1197/jamia.M1535
Clason, D.L., & Dormody, T.J. (1994). Analyzing data measured by individual Likert-
type items. Journal of Agricultural Education, 35(4), 31-35. [PDF format].
Retrieved from
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.113.2197&rep=rep1&t
ype=pdf
Cohen, J. (1992, June). Statistical power analysis. Current Directions in Psychological
Science, 1(3), 98-101. [Electronic source]. Retrieved from
http://journals.sagepub.com/doi/pdf/10.1111/1467-8721.ep10768783
Colgate-Palmolive. (2018). Importance of childhood oral hygiene & the role of parents.
Oral Care for Children. [Electronic source]. Retrieved from
https://www.colgateprofessional.com/education/patient-education/topics/oral-
care-for-children/childhood-oral-hygiene-and-the-role-of-parents
Commonwealth of Kentucky. (2015). Appalachian Regional Commission annual
investment strategy statement: Fiscal year 2015. [PDF format]. Retrieved from
https://www.arc.gov/images/grantsandfunding/StatePlans/KentuckyStrategyState
ment2015.pdf
145
Creswell, J.W. (2014). Research design: Qualitative, quantitative, and mixed methods
approaches (4ed.). Thousand Oaks, CA: SAGE Publications, Inc.
Cullinan, M.P., Ford, P.J., & Seymour, G.J. (2009, September 3). Periodontal disease and
systemic health: Current status. Australian Dental Journal, 54(S1), S62-S69. doi:
10.1111/j.1834-7819.2009.01144.x
Dawkins, E., Michimi, A., Ellis-Griffith, G., Peterson, T., Carter, D., & English, G.
(2013, May 2). Dental caries among children visiting a mobile dental clinic in
south central Kentucky: A pooled cross-sectional study. BioMedical Oral Health,
13(19), 1-9. doi: 10.1186/1472-6831-13-19
Dye, B.A., Vargas, C.M., Lee, J.L., Magder, L., & Tinanoff, N. (2011, February).
Assessing the relationship between children’s oral health status and that of their
mother’s. The Journal of the American Dental Association, 142(2), 173-183. doi:
http://dx.doi.org/10.14219/jada.archive.2011.0061
Elam, C. (2002). Culture, poverty and education in Appalachian Kentucky. Education
and Culture, 18(1), 10-13. [Electronic format]. Retrieved from
http://www.jstor.org/stable/42922461
FAHE. (2015). Appalachian poverty. Our Plan for Appalachia. [Electronic source].
Retrieved from http://fahe.org/
Ferguson, L. (2004). External validity, generalizability, and knowledge utilization.
Journal of Nursing Scholarship, 36(1), 16-22. doi: 10.1111/j.1547-
5069.2004.04006.x
146
Fluoride Action Network. (2016). State fluoride database. Kentucky. [Electronic source].
Retrieved from http://fluoridealert.org/researchers/states/kentucky/
Fluoride Legislative User Information Database. (2012). Kentucky administrative code.
Kentucky. [Electronic source]. Retrieved from
http://fluidlaw.org/policy?tid%5B%5D=109&field_legislation_type_value_many
_to_one%5B%5D=Administrative+Code&keys=&nid_op=%3E%3D&nid%5Bva
lue%5D=-
1&view_name=Advanced_Search_Policy&view_display_id=page_1&view_args
=&view_path=policy&view_base_path=policy&view_dom_id=1&pager_element
=0
Fowler, F.J. (2014). Survey research methods (5th
ed.). Thousand Oaks, CA: SAGE
Publications, Inc.
Frisbee, S.J., Chambers, C.B., Frisbee, J.C., Goodwill, A.G., & Crout, R.J. (2010,
September 18). Self-reported dental hygiene, obesity, and systemic inflammation
in a pediatric rural community cohort. Bio-Med Central Oral Health, 10(21). doi:
10.1186/1472-6831-10-21
Frost, J. (2016, April 6). Best way to analyze Likert item data: Two sample t-test versus
Mann-Whitney. Data Analysis. [Electronic format]. Retrieved from
http://blog.minitab.com/blog/adventures-in-statistics-2/best-way-to-analyze-likert-
item-data:-two-sample-t-test-versus-mann-whitney
Gadermann, A.M., Guhn, M., & Zumbo, B.D. (2012, January). Estimating ordinal
reliability for Likert-type and ordinal item response data: A conceptual, empirical,
147
and practical guide. Practical Assessment, Research & Evaluation, 17(3).
[Electronic source]. Retrieved from https://pareonline.net/pdf/v17n3.pdf
Garcia-Perez, M.A. (2012, August 29). Statistical conclusion validity: Some common
threats and simple remedies. Frontiers in Psychology, 3, 1-11. doi:
10.3389/fpsyg.2012.00325
Glanz, K., Rimer, B.K., Viswanath, K. (Eds.). (2008). Health behavior and health
education: Theory, research, and practice. San Francisco, CA: John Wiley &
Sons, Inc.
Grimm, P. (2010, December 15). Social desirability bias. Wiley International
Encyclopedia of Marketing. doi:
https://doi.org/10.1002/9781444316568.wiem02057
Guo, Y., Logan, H.L., Dodd, V.J., Muller, K.E., Marks, J.G., & Riley, J.L. (2014, July).
Health literacy: A pathway to better oral health. American Journal of Public
Health, 104(7),e85-e91. doi: 10.2105/AJPH.2014.30130
Harrison, J.A., Mullen. P.D., & Green, L.W. (1992). A meta-analysis of studies of the
health belief model with adults. Health Education Research, 7(1), 107-116. doi:
10.1093/her/7.1.107
Hayden, J. (2014). Introduction to health behavior theory (2nd
ed.). Wayne, NJ: Jones &
Bartlett Learning.
Helitzer, D., Hollis, C., Sanders, M., & Roybal, S. (2012). Addressing the ‘other’ health
literacy competencies – knowledge, deispositions, and oral/aural communication:
148
Development of TALKDOC, and intervention assessment tool. Journal of Health
Communications, 17(3), 160-175. doi: 10.1080/10810730.2012.712613
Horowitz, A.M., & Kleinman, D.V. (2008, April). Oral health literacy: The new
imperative to better oral health. Dental Clinics of North America, 52(2), 333-344.
doi: 10.1016/j.cden.2007.12.001
Howlett, B., Rogo, E.J., & Shelton, T.G. (2014). Evidence based practive for health
professionals: An interprofessional approach. Boston, MA: Jones & Bartlett.
Hussain, M. (2012, July). Descriptive statistics – presenting your results. Journal of
Pakistan Medical Association, 62(7), 741-743. [PDF format]. Retrieved from
http://www.jpma.org.pk/PdfDownload/3569.pdf
Hutson, S.P., Dorgan, K.A., Phillips, A.N., & Behringer, B. (2007). The mountains hold
things in: The use of community research review work groups to address cancer
disparities in Appalachia. Oncology Nursing Forum, 34(6), 1133-1139. [PDF
format]. Retrieved from
https://www.etsu.edu/nursing/ccc/documents/hutson_dorgan_phillips_mntsholdthi
ngsin_2008.pdf
IBM Analytics. (n.d.). IBM SPSS software. Technology. [Electronic source]. Retrieved
from http://www.ibm.com/analytics/us/en/technology/spss/
Janz, N.K., & Becker, M.H. (1984). The health belief model: A decade later. Health
Education Quarterly, 11(1), 1-47. [PDF format]. Retrieved from
https://deepblue.lib.umich.edu/bitstream/handle/2027.42/66877/10.1177_1090198
18401100101.pdf
149
Jones, M., Lee, J.Y., & Rozier, R.G. (2007, September). Oral health literacy among adult
patients seeking dental care. The Journal of the American Dental Association,
138(9), 1199-1208. doi: 10.14219/jada.archive.2007.0344
Joseph, G., Burke, N.J., Tuason, N., Barker, J.C., & Pasick, R.J. (2009, October).
Perceived susceptibility to illness and perceived benefits of preventative care: An
exploration of behavioral theory constructs in a transcultural context. Health
Education and Behavior, 36(5 Suppl), 71S-90S. doi: 10.1177/1090198109338915
Kandel, E.A., Richards, J.M., & Binkley, C.J. (2012, September 5). Childhood caries in
the state of Kentucky, USA: A cross-sectional study. BMC Oral Health, 12(38).
[Electronic source]. doi: 10.1186/1472-6831-12-38
Kawashita, Y., Kitamura, M., & Saito, T. (2011, July). Early childhood caries.
International Journal of Dentistry, 2011, 1-7. doi: 10.1155/2011/725320
Kentucky Department of Public Health. Access to dental care. Title V Fact Sheet. [PDF
format]. Retrieved from http://chfs.ky.gov/dph/mch/cfhi/oralhealth.htm
Kentucky Health Facts. (2018). Location: Demographics. Data by Location. [Electronic
source]. Retrieved from http://www.kentuckyhealthfacts.org/data/location/
Kentucky Youth Advocates. (2005). Parents voice concerns about children’s dental care
in their communities. Kentucky’s Cavity. [PDF version]. Retrieved from
http://kyyouth.org/wp-content/uploads/2012/09/Kentuckys-Cavity.pdf
Kentucky Youth Advocates. (2016). Making smiles happen: 2016 oral health study of
Kentucky’s youth. [PDF version]. Retrieved from
150
http://www.kentuckyoralhealth.com/wp-
content/uploads/2016/10/Making_Smiles_Happen_2016.pdf
Kidd, E. (2011, November 2). The implications of the new paradigm of dental caries.
Journal of Dentistry, 39(S2), 1-7. doi: 10.1016/jdent.2011.11.004
Kimberland, C.L., & Winterstein, A.G. (2008, December). Validity and reliability of
measurement instruments used in research. American Journal of Health-System
Pharmacy, 65(1), 2276-2284. doi: 10.2146/ajhp070364
Kostoulas, A. (2014, February, 23). How to interpret ordinal data. Asked and Answered,
Notes, Research Methods. [Electronic source]. Retrieved from
https://achilleaskostoulas.com/2014/02/23/how-to-interpret-ordinal-data
Krause, D.D., May, W.L., Lane, N.M., Cossman, J.S., & Konrad, T.R. (2012, July). An
analysis of oral health disparities and access to services in the Appalachian
region. Appalachian Regional Commission. [Electronic source]. Retrieved from
http://www.arc.gov/research/researchreportdetails.asp?REPORT_ID=100
Laerd Dissertation. (2012a). External validity. Quantitative Dissertations: Research
Strategy. [Electronic source]. Retrieved from
http://dissertation.laerd.com/external-validity.php
Laerd Dissertation. (2012b). Internal validity. Quantitative Dissertations: Research
Strategy. [Electronic source]. Retrieved from
http://dissertation.laerd.com/internal-validity.php
151
Laerd Dissertation. (2012c). Mono-method bias and construct validity. Quantitative
Dissertations: Research Strategy. [Electronic source]. Retrieved from
http://dissertation.laerd.com/construct-validity-p3.php
Laerd Statistics. (2013a). Spearman’s rank-order correlation using SPSS statistics. SPSS
Statistics. [Electronic source]. Retrieved from https://statistics.laerd.com/spss-
tutorials/spearmans-rank-order-correlation-using-spss-statistics.php
Laerd Statistics. (2013b). Mann-Whitney U test using SPSS statistics. SPSS Statistics.
[Electronic source]. Retrieved from https://statistics.laerd.com/spss-
tutorials/mann-whitney-u-test-using-spss-statistics.php
Laerd Statistics. (2013c). Hypothesis testing. The Null and Alternative Hypothesis.
[Electronic source]. Retrieved from https://statistics.laerd.com/statistical-
guides/hypothesis-testing-3.php
Laerd Statistics. (2018). Comparison of distributions. Mann-Whitney U Test. [Electronic
source]. Retrieved from https://statistics.laerd.com/premium/spss/mwut/mann-
whitney-test-in-spss-19.php
Lee, J.Y., Divaris, K., Baker, A.D., Rozier, R.G., & Vann, W.F. (2012, May). The
relationship of oral health literacy and self-efficacy with oral health status and
dental neglect. American Journal of Public Health, 102(5), 923-929. doi:
10.2105/AJPH.2011.300291
Leech, N.L., & Onquegbuzie, A.J. (2002, November). A call for greater use of
nonparametric statistics. Education Resources Information Center. [PDF format].
Retrieved from http://files.eric.ed.gov/fulltext/ED471346.pdf
152
Legislative Research Commission. (2000, August). Adult education and literacy in
Kentucky. Research Report No. 296. [PDF format]. Retrieved from
http://www.lrc.ky.gov/lrcpubs/rr296.pdf
Leicht, K.T. (2013, September). Social change. Oxford Bibliographies. [Electronic
format]. Retrieved from
http://www.oxfordbibliographies.com/view/document/obo-9780199756384/obo-
9780199756384-0047.xml
Ludke, R.L., & Obermiller, P.J. (2016, January 9). Appalachian health and well-being.
University Press of Kentucky. [Electronic source]. Retrieved from
http://www.kentuckypress.com/pages/Appalachian_Health_and_Wellbeing.htm
Ludke, R.L., Obermiller, P.J., Jacobson, C.J., Shaw, T., & Wells, V. (2006). ‘Sometimes
it’s hard to figure’: The functional health literacy of Appalachians in a
metropolitan area. Journal of Appalachian Studies, 12(1), 7-25. [PDF format].
Retrieved from
http://webcentral.uc.edu/eProf/media/attachment/eprofmediafile_1817.pdf
Lund Research Ltd. (2012). Convenience sampling. Laerd Dissertation. [Electronic
source]. Retrieved from http://dissertation.laerd.com/convenience-sampling.php
Madison County Health Department. (2016). School dental health program. Dental
Screenings. [Electronic source]. Retrieved from
http://www.madisoncountyhealthdept.org/SchoolDental.html
McCambridge, J., Willon, J., & Elbourne, D.R. (2014, March). Systematic review of the
Hawthorne effect: New concepts are needed to study research participation
153
effects. Journal of Clinical Epidemiology, 67(3), 267-277. doi:
10.1016/j/jclinepi.2013.08.015
Michael, R.S. (n.d.). Threats to internal & external validity. Indiana University. [PDF
format]. Retrieved from
http://www.indiana.edu/~educy520/sec5982/week_9/520in_ex_validity.pdf
Milfrom, P., Garcia, R.I. Ismail, A., Katz, R.V., & Weintraub, J.A. (2004, October).
Improving America’s access to care: The National Institute of Dental and
Craniofacial Research addresses oral health disparities. The Journal of the
American Dental Association, 135(10), 1389-1396. doi:
10.14219/jada.archive.2004.0048
Miller, E., Lee, J.Y., DeWalt, D.A., & Vann, W.F. (2010, July). Impact of caregiver
literacy on children’s oral health outcomes. Pediatrics, 126(1), 106-114. doi:
10.1542/peds.2009-2887
Milligan, G.W., & McFillen, J.M. (1984, December). Statistical conclusion validity in
experimental designs used in business research. Journal of Business Research,
12(4), 437-462. doi: doi.org/10.1016/0148-2963(84)90024-9
Moser, D.K., Robinson, S., Biddle, M.J., Pelter, M.M., Nesbitt, T.S., Southard, J.,
Cooper, L., & Dracup, K. (2015, August). Health literacy predicts morbidity and
mortality in rural patients with heart failure. Journal of Cardiac Failure, 21(8),
612-618. http://dx.doi.org/10.1016/j.cardfail.2015.04.004
Murray, J. (2013, September). Likert data: What to use, parametric or non-parametric?
International Journal of Business and Social Science, 4(11), 258-264. [Electronic
154
source]. Retrieved from
https://s3.amazonaws.com/academia.edu.documents/45281077/Likert_Data_What
_to_use__Parametric_or_Non-
Parametric.pdf?AWSAccessKeyId=AKIAIWOWYYGZ2Y53UL3A&Expires=15
32966109&Signature=wP0wZWvDZLwlEQ1Q%2B7FoCQ4WzrM%3D&respon
se-content-
disposition=inline%3B%20filename%3DLikert_Data_What_to_Use_Parametric_
or_No.pdf
National Business Research Institute. (2017). Validity of your survey results. Internal
Validity. [Electronic source]. Retrieved from https://www.nbrii.com/our-
process/validity/
National Heart, Lung, and Blood Institute. (2016, June 22). Explore coronary heart
disease risk factors. Health Information for the Public. [Electronic source].
Retrieved from https://www.nhlbi.nih.gov/health/health-topics/topics/hd
National Institute of Dental and Craniofacial Research. (2009, March). A look at oral
health disparities in Appalachia. Interviews with Oral Health Researchers.
[Electronic source]. Retrieved from
http://www.nidcr.nih.gov/research/ResearchResults/InterviewsOHR/COHRA.htm
National Institute of Dental and Craniofacial Research. (2012, August). Seal out tooth
decay. Oral Health. [Electronic source]. Retrieved from
http://www.nidcr.nih.gov/oralhealth/Topics/ToothDecay/SealOutToothDecay.htm
#whatIfA
155
Neiswanger, K., McNeil, D.W., Foxman, B., Goveil, M., Cooper, M.E., Weyant, R.J., . , .
, . , . , . , . , . , . , . , & Marazita, M.L. (2015, May 18). Oral health in a sample of
pregnant women from northern Appalachia (2011-2015). International Journal of
Dentistry, 2015, 1-12. doi: http://dx.doi.org10.1155/2015/469376
Northern Kentucky Health Department. (2016). Dental prevention program. Fact Sheet.
[Electronic source]. Retrieved from http://www.nkyhealth.org/Services/Sealant-
Fact-Sheet.aspx
Nutbeam, D. (2008, December). The evolving concept of health literacy. Social Science
& Medicine, 67(12), 2072-2078. [PDF format]. Retrieved from
http://www.sciencedirect.com/science/article/pii/S0277953608004577
Onwuegbuzie, A.J., & Collins, K.M.T. (2007) A typology of mixed methods sampling
designs in social science research. The Qualitative Report, 12(2), 281-316.
[Electronic format]. Retrieved from http://nsuworks.nova.edu/tqr/vol12/iss2/9/
Oral Health in Kentucky Technical Report. (2016, February). Oral health in Kentucky.
Center for Health Workforce Studies. [PDF format]. Retrieved from
http://chws.albany.edu/archive/uploads/2016/02/Oral_Health_Kentucky_Technica
l_Report_2016.pdf
Paasche-Orlow, M.K., & Wolf, M.S. (2007, August). The causal pathways linking health
literacy to health outcomes. American Journal of Health Behavior, 31(S1), S19-
S26. [Electronic format]. Retrieved from
http://www.ingentaconnect.com/content/png/ajhb/2007/00000031/A00100s1/art0
0004
156
Pandya, M., & Desai, C. (2013,). Compensation in clinical research: The debate
continues. Perspectives in Clinical Research, 4(1), 70-74. doi: 10.4103/2229-
3485.106394
Patrick, M., & Thomas, T. (2016, October 28). Study: Kentucky kids have more access to
oral health care but still have poor outcome. Fluoride Action Network. [Electronic
source]. Retrieved from http://fluoridealert.org/news/study-kentucky-kids-have-
more-access-to-oral-health-care-but-still-have-poor-outcome/
Peterson-Sweeney, K., & Stevens, J. (2010, August). Optimizing the health of infants and
children: Their oral health counts! Journal of Pediatric Nursing, 25(4), 244-249.
doi: 10.1016/j.pedn.2009.03.003
Polk, D.E., Weyant, R.J., Crout, R.J., McNeil, D.W., Tarter, R.E., Thomas, J.G., &
Marazita, M.L. (2008). Study protocol of the Center for Oral Health Research in
Appalachia etiology study. BMC Oral Health, 8(18). [Electronic format].
Retrieved from http://bmcoralhealth.biomedcentral.com/articles/10.1186/1472-
6831-8-18
Pollard, K., & Jacobsen, L.A. (2011, September). The Appalachian region in 2010: A
census data overview. Appalachian Regional Commission. [PDF format].
Retrieved from http://www.prb.org/pdf12/appalachia-census-chartbook-2011.pdf
Privitera, G.J. (2011). Statistics for the behavioral sciences. Los Angeles, CA: Sage
Publications, Inc.
Reed, P. (2016, May 18). Sealant programs critical to Kentucky’s oral health turnaround.
Inside Oral Healthcare: Kentucky Educational Television. [Electronic format].
157
Retrieved from http://www.ket.org/oral-health/sealant-programs-critical-to-
kentuckys-oral-health-turnaround/
Reisi, M., Mostafavi, F., Javadzade, H., Mahaki, B., Tavassoli, E., & Sharifirad, G.
(2016, January). Impact of health literacy, self-efficacy, and outcome expectations
on adherence to self-care behaviors in Iranians with type 2 diabetes. Oman
Medical Journal, 31(1), 52-59. doi: 10.5001/omj.2016.10
Research Methodology. (2018). Convenience sampling. Sampling. [Electronic source].
Retrieved from https://research-methodology.net/sampling-in-primary-data-
collection/convenience-sampling/
Rhee, K. (2008, January). Childhood overweight and the relationship between parent
behaviors, parenting style, and family functioning. The Annals of the American
Academy of Political and Social Science, 615(1), 11-37. doi:
10.1177/0002716207308400
Rosenstock, I.M. (1974, December). Historical origins of the health belief model. Health
Education & Behavior, 2(4), 328-335. [Electronic format]. Retrieved from
http://heb.sagepub.com.ezproxy.uky.edu/content/2/4/328.full.pdf+html
Rosenstock, I.M., Strecher, V.J., & Becker, M.H. (1988). Social learning theory and the
Health Belief Model. Health Education & Behavior, 15(2), 175-183. [Electronic
format]. Retrieved from
http://heb.sagepub.com.ezproxy.uky.edu/content/15/2/175.full.pdf+html
Rubin, G.J., Bakhshi, S., Amlot, R., Fear, N., Potts, H.W.W., & Michie, S. (2014,
November). The design of a survey questionnaire to measure perceptions and
158
behavior during an influenza pandemic: the flu telephone survey template
(FluTEST). Health Services and Delivery Research, 2, (41). doi:
10.3310/hsdr02410
Saman, D.M., Johnson, A.O., Arevalo, O., & Odoi, A. (2011). Geospatially illustrating
regional-based oral health disparities in Kentucky. Public Health Reports, 126(4),
612-618. [Electronic format]. Retrieved from
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3115226/
Saris, W.E., & Gallhofer, I.N. (2007). Design, evaluation, and analysis of questionnaires
for survey research. Hoboken, NJ: John Wiley & Sons, Inc.
Savage, M.W., Scott, A.M., Aalboe, J.A., Stein, P.S., & Mullins, R. (2015. February 1).
Perceptions of oral health in Appalachian Kentucky: Implications for message
design. Health Communications, 30, 186-195. doi:
10.1080/10410236.2014.974127
Schwarzer, R., Schuz, B., Ziegelmann, J.P., Lippke, S., Luszcynaska, A., & Scholz, U.
(2007, June). Adoption and maintenance of four health behaviors: Theory-guided
longitudinal studies on dental flossing, seat belt use, dietary behavior, and
physical activity. Annals of Behavioral Medicine, 33(2), 156-166. doi:
10.1007/BF02879897
Schunk, D.H., & Pajares, F. (2009). Self-efficacy theory. In K.R. Wentzel & A. Wigfield
(Eds.), Handbook of Motivation at School (35-53). New York, NY: Routledge.
Scott, T.A. (n.d.). Sample size planning, calculation, and justification. Vanderbilt
University Department of Biostatistics. [PowerPoint format]. Retrieved from
159
http://biostat.mc.vanderbilt.edu/wiki/pub/Main/TheresaScott/SampleSize.TAScott
.handout.pdf
Sijtsma, K. (2009, March). On the use, the misuse, and the very limited usefulness of
Cronbach’s alpha. Psychometrika, 74(107), 107-120. doi: 10.1007/s11336-008-
9101-0
Sischo, L., & Broder, H.L. (2011, January 18). Oral health-related quality of life: What,
why, how, and future implications. Journal of Dental Research, 90(11), 1264-
1270. doi: 10.1177/0022034511399918
Smith, S. (2013, April). Determining sample size: How to ensure you get the correct
sample size. Qualtrics. [Electronic format]. Retrieved from
https://www.qualtrics.com/blog/determining-sample-size/
Statistics Solutions. (2017). Correlation (Pearson, Kendall, Spearman). Directory of
Statistical Analyses. [Electronic source]. Retrieved from
http://www.statisticssolutions.com/correlation-pearson-kendall-spearman/
Steckler, A., & McLeroy, K.R. (2008, January). The importance of external validity.
American Journal of Public Health, 98(1), 9-10. doi: 10.2105/AJPH.2007.126847
Strauss, M.E., & Smith, G.T. (2009). Construct validity: Advances in theory and
methodology. Annual Review of Clinical Psychology, 5(1), 1-25. doi:
10.1146/annurev.clinpsy.032408.153639
Strecher, V.J., & Rosenstock, I.M. (1997). The health belief model. In Baum, A.,
Newman, S., Weinman,, J., West, R., & McManus, C (Eds.), Cambridge
160
handbook of psychology, health, and medicine (113-121). Cambridge, United
Kingdom: Cambridge University Press.
Sullivan, G.M., & Artino, A.R. (2013, December). Analyzing and interpreting data from
Likert-type scales. Journal of Graduate Medical Education, 5(4), 541-542. doi:
10.4300/JGME-5-4-18
Sullivan, G.M., & Feinn, R. (2012, September 4). Using effect size – or why the P value
is not enough. Journal of Graduate Medical Education, 4(3), 279-282. doi:
10.4300/JGME-D-12-00156.1
Talekar, B.S., Rozier, R.G., Slade, G.D., & Ennett, S.T. (2005, March). Parental
perceptions of their preschool-aged children’s oral health. The Journal of the
American Dental Association, 136(3), 364-372. doi:
http://dx.doi.org/10.14219/jada.archive.2005.0179
Tavakol, M., & Dennick, R. (2011). Making sense of Cronbach’s alpha. International
Journal of Medical Education, 2, 53-55. doi: 10.5116/ijme.4dfb.8dfd
Trochim, W.M.K. (2006a, October). Likert scaling. Research Methods Knowledge Base.
[Electronic source]. Retrieved from
http://www.socialresearchmethods.net/kb/scallik.php
Trochim, W.M.K. (2006b, October). Construct validity. Research Methods Knowledge
Base. [Electronic source]. Retrieved from
http://www.socialresearchmethods.net/kb/constval.php
Trochim, W.M.K. (2006c, October). Measurement validity types. Research Methods
Knowledge Base. [Electronic source]. Retrieved from University of Surry. (2001).
161
The importance of pilot studies. Social Research Update, 35. [Electronic source].
Retrieved from http://sru.soc.surrey.ac.uk/SRU35.html
Trochim, W.M.K. (2006d, October). External validity. Research Methods Knowledge
Base. [Electronic source]. Retrieved from
https://www.socialresearchmethods.net/kb/external.php
Trochim, W.M.K. (2006e, October). Internal validity. Research Methods Knowledge
Base. [Electronic source]. Retrieved from
https://www.socialresearchmethods.net/kb/intval.php
United States Census 2010. (n.d.). 2010 census population profile maps. 2010 Census
Data. [Electronic source]. Retrieved from
http://www.census.gov/2010census/data/
University of Kentucky Dentistry. (2012). Dental outreach programs. Outreach and
Services. [Electronic source]. Retrieved from https://dentistry.uky.edu/dental-
outreach-programs
University of Leicester. (2017). Measures of variability: the range, inter-quartile range
and standard deviation. Numerical Data. [Electronic source]. Retrieved from
http://www2.le.ac.uk/offices/ld/resources/numerical-data/variability
University of Minnesota. (2017). Threats to the validity of your design. Interactive
Learning Modules. [Electronic source]. Retrieved from
https://cyfar.org/ilm_3_threats
162
University of Pittsburg. (2016). COHRA1 cohort study. School of Dental Medicine.
[Electronic source]. Retrieved from http://www.dental.pitt.edu/cohra1-cohort-
study
University of Tennessee. (2017). Informed consent. Office of Research Integrity.
[Electronic source]. Retrieved from https://www.utc.edu/research-
integrity/institutional-review-board/informedconsent/
University of Virginia Library. (2017). Using and interpreting Cronbach’s alpha.
Research Data Services + Sciences. [Electronic format]. Retrieved from
http://data.library.virginia.edu/using-and-interpreting-cronbachs-alpha/
University of Wisconsin. (2017). Types of quantitative research. Study Designs.
[Electronic source]. Retrieved from
http://researchguides.ebling.library.wisc.edu/c.php?g=293229&p=1953454
U.S. Department of Health & Human Services. (2016, March 16). Informed consent
checklist. Office for Human Research Protections. [Electronic format]. Retrieved
from https://www.hhs.gov/ohrp/regulations-and-
policy/guidance/checklists/index.html
U.S. Department of Health and Human Services. (2018). Quick guide to health literacy.
Office of Disease Prevention and Health Prevention. [PDF version]. Retrieved
from https://health.gov/communication/literacy/quickguide/factsbasic.htm
U.S. Food and Drug Administration. (2016, January 25). Payment to research subjects –
information sheet. Guidance for Institutional Review Boards and Clinical
163
Investigators. [Electronic source]. Retrieved from
https://www.fda.gov/RegulatoryInformation/Guidances/ucm126429.htm
Van der Weijden, F., & Slot, D.E. (2010, December 7). Oral hygiene in the prevention of
periodontal diseases: The evidence. Periodontology 2000, 55(1), 104-123. doi:
10.1111/j.1600-0757.2009.00337.x
Vann, W.F., Lee, J.Y., Baker. D., & Divaris, K. (2010). Oral health literacy among
female caregiver’s: Impact on oral health outcomes in early childhood. Journal of
Dental Research, 89(12), 1395-1400. doi: 10.1177/0022034510379601
Warmbrod, J.R. (2014). Reporting and interpreting scores derived from Likert-type
scales. Journal of Agricultural Education, 55(5), 30-47. doi:
10.5032/jae.2014.05030
Wyoming Department of Health. (2018). Prevention vs. treatment. What is Prevention?
[Electronic source]. Retrieved from
https://health.wyo.gov/publichealth/prevention/substanceabuseandsuicide/what-is-
prevention/prevention-vs-treatment/
Xu, X.Y., & Leung, A.Y.M. (2018). Health literacy, self-efficacy, and associated factors
among patients with diabetes. Health Literacy Research and Practice, 2(2), 67-
77. doi: 10.3928/24748307-20180313-01
164
Appendix
Individual survey items used in this study.
I know that I am volunteering in a research study.
If parents have poor teeth, their kids will too.
I floss my teeth every day.
Having cavities is a normal part of life.
Sometimes I am just too busy to brush my teeth twice a day.
Having poor teeth can lead to other health problems.
My kids floss their teeth every day.
If I brush and floss every day, I will have good teeth.
My kids have had cavities in their teeth.
Sometimes my kids don't brush their teeth twice a day.
People are born with either good teeth or bad teeth and they stay that way for their whole life.
I have had cavities in my teeth.
I brush my teeth twice every day.
Sometimes I am just too busy to floss my teeth every day.
Sometimes having poor teeth just happens and people can't help that.
I can only have good teeth if I see a dentist a lot.
If kids brush their teeth every day while they are young, they will have better teeth as they age.
If kids floss their teeth every day while they are young, they will have better teeth as they age.
As an adult I can still have poor teeth even if I brush and floss a lot, so why bother.
Kids can only have good teeth if they see a dentist a lot.
Sometimes my kids don't floss their teeth every day.
My kids brush their teeth twice every day.
If someone has bad teeth, they can only get better by seeing a dentist.
Having poor teeth can make me sick in other ways.
My yearly household income can be described as:
Below $31,300 Above $31,300
My education level can be described as:
Less than a high school degree High school degree or more